Full Citation

Transcription

Full Citation

International Journal of Information Technology and Computer Science ( IJITCS ) ( ISSN : 2091 – 1610 )
Analysis of DQT and DHT in JPEG Files
Nor Amira Nor Azhan
Department of Computer System and Communications
Faculty of Computing, Universiti Teknologi Malaysia
Skudai, Johor Bahru
[email protected]
Ikuesan R. Adeyemi
Skudai, Johor Bahru
[email protected]
Shukor Abd Razak
Skudai, Johor Bahru
[email protected]
Abstract :
JPEG files are the most popular image format that is commonly analysed in image forensics research areas.
The compositions of these files are prominent to help accomplish the problem of source identification, content
classification, forgery detection, steganography, encryption, and file recovery. The purpose of this paper is to
present the overview of JPEG file format, as well as identifying important characteristics of its content, which
are related to signature and compression features. For analysis purposes, we used about 80 JPEG files
produced from several digital devices to examine their contents. The results show that it is possible to
differentiate images among JPEG files and from other file types, by examining specific content of JPEG files.
Keywords-component; JPEG file; signature; compression
I. Introduction
JPEG file standard, a file developed by the Joint Photographic Expert Group, is the commonly used
image file format found in most electronic and storage devices [2-6]. Additionally, JPEG files are usually the
default image format in almost all of digital cameras. Statistics reveals by global camera industries revealed that
over $55 billion digital camera sales on 2010 , while the figure is projected to increase to over $65 billion by
2015 [7], thus making JPEG files the order of the digital world of photography.
80% to 90% of the criminal cases today have some kind of digital evidence [8] and one of such
evidence is digital image file. Details analysis of JPEG content is critically required for image forensic research
areas, since it is the widely used image format. The process of determining the device-source of an image is
called image source identification. Image content classification and forgery detection are two sides of the
forensic process. Image content classification analyzes image files for specific purpose, such as cyber
pornography. On the other hand, image forgery detection field determines whether the images have been
manipulated and altered. Image steganography is a technique that can be used to hide secret information in
This papers was presented at 2
nd
International Conference on E-Education & Learning Technologies ( ICEELT 2013 ) ,
Singapore on May 4 – 5 , 2013 ……. Page … 1……
image files while image encryption is a technique used to prevent the possible interpretation of an image file, so
as to render such file unreadable to unintended audience. However, file recovery methods are needed to recover
JPEG files that has been deleted or corrupted. Without the understanding of the underlying file structure, and bit
positioning, it is almost impossible to recover a deleted or corrupted file irrespective of total or partial deletion
process. This proves that the detail studies of JPEG file inner content’s is necessary in order to help and solve
specific research problems.
This paper discusses the basic information of JPEG files with special consideration on deep analysis of
their internal structure. The next section begins with the discussions of basic information of JPEG files, their
content and available tools that can be used with this file type. Furthermore, the paper details experiment and
result of this process in Section 3, clarifying the important component contained in JPEG files. Concise
conclusion elaborating current research challenges related to JPEG file image content analysis is also presented.
II. Background of Study
JPEG was created as a standard format to store photographic images. It only specifies how component
values are stored but not the way colours need to be represented. JPEG File Interchange Format (JFIF), another
acronym for JPEG, was introduced to allow images to be created and exchanged among different applications
[9]. Usually, JPEG file has the extension of .jpg, .JPG, .jpeg or .JPEG.
Generally, JPEG file uses lossy compression methods that excludes some bits string (which is not
detectable by the human eyes), before images are stored so as to minimize storage space as well as reduce file
transmission time. For instance, to store 1 MB Windows BMP file, it requires only 50 KB with JPEG file [2].
There are four steps involve in compressing (or encoding) the JPEG file which are sampling, Discrete Cosine
Transform, quantization and entropy coding. The reverse process is also true -decompression (or decoding).
Miano (1999) elaborates more on Compressed Image File Format.
JPEG files are defined by four compression modes which are; hierarchical, progressive, sequential, and
lossless. For file interchange process, the sequential compression method with Huffman baseline encoding [10]
is strongly recommended to be used [9]. In sequential mode, images are encoded from top to bottom.
Throughout this paper, discussions will focus more on baseline JPEG file.
A. State of the Art of JPEG File
The specification of JPEG file format or specifically markers can be in three different formats, which
are JFIF (Figure 1), JFIF Exif (Figure 2) Exif (Figure 3). JFIF Exif is non-standard format in which both JFIF
and Exif markers can be found in one image file [11].
nd
SOI
SOI
APP₀
APP₀
*APP₀ Extension
*APP₀ Extension
DQT
DQT
SOF
SOF
DHT
DRI
SOS
DHT
Scan
SOS 1
EOI
Scan 1
SOI
*Thumbnail
SOI
APP₁
SOI
APP₀
*APP₂
DQT
Exif
DQT
DHT
EOF
DHT
SOF
*DRI
SOS
SOF
Scan
SOS
EOI
Figure 1(a)
DRI
Scan
DHT
EOI
SOS n
Figure 2. JFIF Exif Format
*optional
Figure 3. Exif Format [1]
Scan n
EOI
*optional
Figure 1(b). [2]
Figure 1. JFIF Format
Figure 1, 2 and 3 shows JPEG markers contained in the file. Markers are used to specify the location of
component in a file. ‘Scan’ in the figures signifies a compressed data in the file. JFIF format can be identified
with APP₀ marker as shown in Figure 1(b). However, most of the JFIF file do not use the DRI marker, thus the
format in Figure 1(a) is often used.
Exif format can be identified with APP₁ marker and sometimes with APP₂ marker (FlashPix extension
data). Most of the Exif files are created with thumbnail image, usually without the DRI marker as shown in
Figure 3. On the other hand, the JFIF Exif format contains two or three markers, APP₀, APP₁ and APP₂ which
are incompatible as shown in Figure 2. In this case, the APP₀ marker comes without their extension and use the
same Exif format. This specification is not realized by many applications and therefore working with this format
might be a bit difficult.
Exif specifications allow one or two thumbnails to be embedded in the JPEG file [12]. Exif image
specification is described in Figure 4. It shows that the locations of DQT (Define Quantization Table) and DHT
(Define Huffman Table), are may be switched with each other depending on the devices that produced such
image.
nd
Figure 4. Image with Thumbnail/s Specifications
B. Tools
Exif format contain metadata or Exif data that describes the digital image itself. Basically, information
like camera model used, creation time and date are usable for investigation purposes. For the sake of metadata
viewer, there are several available tools that can be used as provided in Graphic Software web site [13].
Additionally, there are also approaches on how to identify only the JFIF header based on their markers [14-16].
Besides that, WinHex [17] and Hexeditor [18] are other tools that displays the actual content of JPEG files in
hexadecimal form.
III. Analysis, Result and Discussions of JPEG Tables
DQT and DHT are important tables in JPEG files that allows the process of compression and
decompression. There are several works done to display the contents of DQT and DHT. However, no significant
exploration have been presented partaining to the analyses of JPEG Metadata. JPEG Metadata Viewer [11] and
JPEGsnoop [19] as open source tool that can be used for such purpose. These softwares synoptically displays
the common information about the JPEG tables including markers, length of data and marker locations in a file.
This section will start with introduction of DQT and DHT, followed by examination of their contents.
For analysis purposes, we used total of 80 JPEG files, in which each 40 files are generated from 20 digital
cameras and mobile cameras. Each file was renamed from 1 to 80 names and the list of their associated devices
is shown in Table 1. The files are image collection that has not undergone any modification. However, two
images of every four images from each device (in no particular sequence), are rotated in order to obtain the
proper display of images. In some cases, these rotated images exhibits some variance with the original image.
This will be discussed further in next sub sections.
TABLE I.
DISTRIBUTION OF DEVICES USED IN EXPERIMENT
Digital Cameras
Device
Mobile Devices
File Name
Device
(.JPG/.jpg)
File Name
(.JPG/.jpg)
1.
Canon EOS 450D
1-4
11. Samsung GT-I9300
41-44
2.
Nikon Coolpix L16
5-8
12. Sony Ericsson K770i
45-48
nd
3.
Nikon D90
9-12
13. Sony Ericsson K850i
49-52
4.
Olympus T105, T100, X36
13-16
14. Sony Ericsson W660i
53-56
5.
Panasonic DMC-F2
17-20
15. Sony Ericsson W705
57-60
6.
Ricoh R50
21-24
16. LG Electronics
61-64
7.
Samsung ES60/SL105/ES63
25-28
8.
Samsung S630
29-32
17. Nokia N73
69-72
9.
Sony Cybershot
33-36
18. Nokia 1234
73-76
37-40
19. Nokia 5320
77-80
10. Sony DSC-W570
GW520
65-68
20. Nokia 6233
A. DQT
Quantization is a sort of matrices used by JPEG to control image compression. Every digital device and
graphical application that produces JPEG files will have their own quantization table as implied in device
setting. For instance, image can be set by low, medium or high quality and this mean users can choose the
quality level to save the image.
Figure 5. Image Pixels
Image is a collection of information about very small dots or represented as pixel. In the first step of
JPEG encoding, each pixel data is grouped into 8x8 blocks as shown in Figure 5. Each component is then
converted from RGB (Red, Green, Blue) to YCbCr (Y: luminance, CbCr: chrominance) colorspace. Luminance
represent brightness of the pixels, chrominance Cb specified blueness while chrominance Cr specified redness.
In the second step of JPEG encoding, DCT is used to transform each 8x8 block into a set of 64 DCT
coefficients. The first value represents the DC coefficient and remaining 63 values are called AC coefficients.
The third phase of JPEG encoding is to quantize the DCT coefficients by using 64 quantization values provided
in DQT. Thus, the results will be used in the next encoding phase, which is entropy encoding.
Figure 6 shows the parameter specification of DQT. Basically, the value of length is calculated start
with length field until the end of quantization values. From 40 images together with the 46 thumbnails of digital
cameras, we found that 67 of them use 67-byte size of DQT, 54 of them use 132-byte size of DQT, and 4 of
them use 197-byte size of DQT. From 40 images together with the 40 thumbnails of mobile cameras, we found
that 80 of them use 67-byte size of DQT and 36 of them use 132-byte size of DQT. It shows that majority of the
images use 67-byte size of DQT. A luminance table can be identified by 00 value while chrominance with 01
value. The quantization values are stored in zigzag order in JPEG files. Hexeditor tool is used to view the
content of DQTs of one selected image as shown in Figure 7. Each DQT starts with FFDB marker and it shows
that this image contains two DQTs.
nd
Marker
Length
Y or CbCr Table
Quantization Values
FF DB
00 43
00 or 01
..........
2-byte
2-byte
1-byte
64-byte
Figure 6. DQT Specification
Figure 7. DQTs in JPEG File
The actual reordering of quantization values are obtained by converting the zigzag order as shown in Figure 8
by using JPEGsnoop [19].
Figure 8. Actual Ordering of Quantization Values
Refer to Figure 8, the precisions 8 bit represent the 8-bit per pixel sample luminance and chrominance.
Quantization tables consist of luminance and chrominance matrices value. This means that the value of image
brightness, blueness and redness have their compression ratio. JPEG may consists of two (one for Y, one for
CbCr) or three DQTs (each for Y, Cb, Cr). Basically, luminance table become the first location of DQT, but
there is a case where it becomes the second location of DQT. Table 2 shows the number of DQTs in 80 JPEG
files.
nd
TABLE II.
Image
NUMBER OF DQTS IN JPEG FILES
#Camera
#Mobile
Images
Images
YCbCr
2
3
Y with default CbCr
-
1
14
16
-
-
-
5
16
11
-
-
3
2
1
-
4
2
-
-
-
-
40
40
Image with 1
Image: YCbCr
Thumbnail
Thumbnail: YCbCr
Image: YCbCr
Thumbnail: Y with default CbCr
Image: Y with default CbCr
Thumbnail: YCbCr
Thumbnail: Y with default CbCr
Image with 2
Image: YCbCr
Thumbnails
1st Thumbnail: YCbCr
nd
2 Thumbnail: YCbCr
Image: YCbCr
nd
2 Thumbnail: Y with default CbCr
Image: YCbCr
1st Thumbnail: Y with default CbCr
2nd Thumbnail: Y with default CbCr
nd
2 Thumbnail: Y with default CbCr
nd
2 Thumbnail: YCbCr
2nd Thumbnail: YCbCr
Total:
Result in Table 2 shows that only five images do not have thumbnail and the rest of images may
contain either one or two thumbnails. Majority of images (14 images from cameras, 16 from mobile) are
provided with luminance and chrominance DQT tables. In the second highest number of images (16 images
from cameras, 11 from mobile), whenever luminance table is provided, the default chrominance table is used.
The default table values are different depending on the devices and thus give the various quality levels of
images.
In this study, JPEG was saved at 90%, with 10% of the data has loss. Although the quality values has
been identified for several digital cameras [20], the quantization table in images may not match that quality
values. Therefore, the last quality saved in the JPEG can be identified by approximating the quality value based
on DQT. The following figures shows the result of approximated quality value for 80 JPEG files by using a
method developed by Neal Krawetz [21].
nd
Average Quality (%)
Digital Camera
100
98
98
98
97
95
93
98
88
90
82
80
70
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39
File Name (.JPG/.jpg)
Average Quality (%)
Figure 9. Approximated Quality Value for Digital Camera Images
Mobile Camera
150
100
50
0
96
83 77 94 81 86
88
76
41 43 45 47 49 51 53 55 57 59 61 63 65 67 69 71 73 75 77 79
File Name (.JPG/.jpg)
Figure 10. Approximated Quality Value for Mobile Camera Images
Most of mobile devices produce the same quality value for images compared to digital cameras with
different quality level. This is due to various quality setting provided by the cameras. The average quality value
is quite high for most images that generated from digital cameras compared to mobile cameras. Most of the
images have the quality value more than 90% while below 90% for the mobile cameras. This shows that images
produced mobile camera will throw out more information and thus the quality of the images are lower.
Results from experiment shows that even if DQT is missing and there is requirement to view the
images, it is possible to use DQT from other available image [4]. As a result, the quality of the image will be
different, either higher or lower, but image still can be viewed. However, even if one byte data from DQT is
modified and changed, the image become corrupted file and cannot be viewed. This type of error can be found
in one of dataset available in the Internet [22], in which one byte of DQT length field has been changed in a
JPEG file.
B. DHT
Other mechanism to control JPEG compression is entropy encoding. There are two standard ways to
compress JPEG file either by using Huffman encoding or arithmetic coding. DHT is important to obtain
successful decoding and it is provided to eliminate the zero numbers in 64 quantized coefficients by applying
variable length codes. Thus, shorter codes will be assigned to more frequently used bits. Figure 11 shows the
parameter specification of DHT.
Marker
DHT Length
Table Class
nd
Table Destination
Huffman Code
Values of Huffman code
FF C4
..........
0 or 1
0 or 1
……….
……….
2-byte
2-byte
4-bit
4-bit
16-byte
No fix length
Figure 11. DHT Specification
Refer to Figure 11, again, the value of length is calculated start with length field until the end of Huffman
code values. From 80 images being analyzed, it shows that all of the Y-DC table use either 31-byte or 418-byte
size of DHT, Y-AC and CbCr-AC table use 181-byte size of DHT, CbCr-DC table use 31-byte size of DHT.
Hexeditor tool is used to view the content of DHTs of one selected image as shown in Figure 12. Each DHT
in that image starts with FFC4 marker and thus this image contains four DHTs.
Figure 12. DHTs in JPEG File
There are four DHT tables that specified in table class and definition field. They are Y-DC, Y-AC, CbCr-DC
and CbCr-AC tables. DQTs and DHTs are related with each other in order to further compress the data.
The observation shows that:
-
If either images or thumbnails are using DQT of Y with default CbCr, then they are using DHT of 2
DC tables
-
If either images or thumbnails are using DQT of YCbCr, then they are using DHT of 2 DC and 2 AC
tables
Refer to Table 2, it shows that there are higher number of images that use DQT of Y with default CbCr (27
images) and DQT of YCbCr (30 images). Thus, the two situations described above are quite important, since a
total of 57 out of 80 images matches with this situation. As a result, there are also higher number of images that
consist of 2 DC tables, and images with 2 DC and AC tables. On the other hand, emphasizing on JFIF files, 100
of this type of files uses 2 DC and AC tables [23]. Additionally, emphasizing on Exif files, 69 out of 76 files
uses 2 DC tables [24].
IV. Discussion
There are several tools/approaches for viewing the important components (DQT and DHT) in JPEG
file, from which valuable forensic information can be extracted. Considering DQT for instance, the image
quality value can be estimated, while number of information about the used device can also be revealed. The
analysis presented in this paper, shows that some devices produce similar image quality value. However, they
also exhibit some characteristics which can be used to highlight the default quality level of the device. If the
default quality levels for the images are known, then the number of resaved image can be determined. This
nd
findings therefore provides useful insight for image modification analysis, as well as to image source
identification.
As shown in the analysis in section 3, the number of DQT is related to DHT. Thus, images can be
concluded to use 2 DC tables if they are using DQT of Y with default CbCr; otherwise, 2 DC and 2 AC tables, if
they use DQT of CbCr. Thus, from this result, the length of the DHT can be estimated to depend on the DQTs
that constitute the image. However, the length cannot be estimated with accurate number because the Huffman
code values are not fixed. However, with this study of specific format of DQT and DHT, it is useful in
differentiating JPEG file from other file types.
Additionally, in comparison with previous research works, this research focuses on one type of JPEG
file comprising various JPEG formats such as Exif and JFIF Exif. It is worthwhile to mention that the JFIF Exif
format is not recognized as standard JPEG format. However, this format is usually ignored by most of the
research work, while we found that most of the rotated images used this type of format. From the analysis, we
found that, due to the effect of image rotation, for both digital and mobile cameras, the un-rotated image
(original image form) uses Exif specification, while rotated images uses JFIF Exif specification.
V. Conclusion
This paper discussed an overview of JPEG file format, as well as identifying important characteristic of
JPEG files. While being an important component for compression and decompression, valuable information
(evidence) can also be extracted from DQT and DHT. We used about 40 JPEG files from digital cameras and 40
JPEG files from mobile cameras to examine the image content. From the examination of DQT, the
approximated quality value of images can be calculated depending on various devices that are used to produce
JPEG files. On the other hand, the relation between DQT and DHT can be used to differentiate between JPEG
files from other file types as well as differentiating between JPEG files themselves.
REFERENCES
[1]
JEIDA, "Digital Still Camera Image File Format Standard (Exif) Version 1.0," 1995.
[2]
J. Miano, Compressed Image File Formats: JPEG, PNG, GIF, XBM, BMP: Association for Computing
Machinery, Inc. (ACM), 1999.
[3]
M. I. Cohen, "Title," unpublished|.
[4]
H. T. Sencar and N. Memon, "Identification and Recovery of JPEG Files with Missing Fragments,"
presented at the Digital Forensic Research Workshop, 2009.
[5]
M. Xu and S. Dong. 2009, Reassembling the Fragmented JPEG Images Based on Sequential Pixel
Prediction.
[6]
J. v. d. Bos and T. v. d. Storm, "Bringing Domain-Specific Languages to Digital Forensics," ICSE ’11,
2011.
[7]
reportlinker.com. 2012, Market Research Reports. Available:
http://www.reportlinker.com/ci02061/Camera.html
[8]
K. Medaris and R. Mislan. 2008, Expert: Digital evidence just as important as dna in solving crimes.
Available: http://news.uns.purdue.edu/x/2008a/080425T-MislanPhones.html
[9]
E. Hamilton, "JPEG File Interchange Format Version 1.02 " 1992.
nd
[10] D. A. Huffman, "A Method for the Construction of Minimum-Redundancy Codes," Proceedings of the
I.R.E, pp. 1098–1102, 1952.
[11] K. M. Mohamad and M. M. Deris, "Visualization of JPEG Metadata," IVIC 2009, LNCS 5857, pp. 543–
550, 2009.
[12] M. M. Deris, et al., "Carving JPEG Images and Thumbnails Using Image Pattern Matching," 2011 IEEE
Symposium on Computers & Informatics, pp. 78-83, 2011.
[13] About.com. 2013, Graphics Software. Available:
http://graphicssoft.about.com/od/exifsoftware/EXIF_IPTC_XMP
[14] K. M. Mohamad and M. M. Deris, "Single-byte-marker for Detecting JPEG JFIF Header using
FORIMAGE-JPEG," 2009 Fifth International Joint Conference on INC, IMS and IDC, pp. 1693-1698,
2009.
[15] K. M. Mohamad, et al., "Dual-Byte-Marker Algorithm for Detecting JFIF Header," ISA 2010, CCIS 76,
pp. 17 – 26, 2010.
[16] K. M. Mohamad, et al., "Detecting JFIF Header using FORHEADER," International Journal of Security
and Its Applications, vol. 5, pp. 23-36, 2011.
[17] X.-W. S. T. AG, "WinHex," 2012.
[18] M. Horz, "Hexeditor," 2009.
[19] C. Hass. 2012, JPEGsnoop. Available: www.impulseadventure.com/photo/
[20] C. Hass. 2008, Comparing JPEG Quality. Available: http://www.impulseadventure.com/photo/jpegquality.html
[21] N. Krawetz. 2005, jpegquality.c. Available: http://www.hackerfactor.com/src/jpegquality.c
[22] B. Carrier. 2005, Digital Forensics Tool Testing Images Available: http://dftt.sourceforge.net/
[23] K. M. Mohamad and M. M. Deris. 2009, Fragmentation Point Detection of JPEG Images at DHT Using
Validator. 173-180.
[24] M. Karresand and N. Shahmehri, "Reassembly of Fragmented JPEG Images Containing Restart Markers,"
presented at the European Conference on Computer Network Defense, 2008.
nd
An Analysis of User Requirements for Virtual Classroom/Collaboration Software
Jin Tan
Information Services
Cardiff University, Cardiff, UK
[email protected]
Abstract:
This article aims to present an analysis of user requirements for virtual classroom/collaboration
software (VCCS) in learning and teaching. It collates and priorities existing user requirements in Cardiff
University, reviews tools in the market, and recommends good tools that meet these requirements.
Keywords-component: virtual classroom, user requirements, learning activities
I. Introduction
Virtual Classroom is not a completely innovate term to educators. In the early 1990s, educators have
started to see the value of computer-mediated communication (CMC), for instance, Hiltz and Turoff [1] described
communication structures in their designed online environment for teaching and learning, named EIES 2. This
could be seen as an example of the prototype of virtual classroom using in distance education. In a recent study
conducted by McBrien, Jones and Cheng [2], it discussed student-perceived strengths and weakness of using a
synchronous online learning platform - Elluminate Live1. The study concluded that most students had a positive
online experience using synchronous learning technology but it closely related to how the instructors control
communication and interaction over online classes on both technological aspects and pedagogical aspects.
It is hardly doubt that with the rapid improvement of information technologies, new trends continue to
impact on the higher education. For example, Howell, Williams and Lindsay [3] stated six trends: (1) the
increasing percentage of adult learners and international students, (2) institutions are becoming more learnercentered, non-linear, and self-directed, (3) a huge growth in Internet usage andtechnological fluency,(4) lifelong
learning is becoming a competitive necessity, (5) more courses, degrees, and universities are becoming available
through distance-education programs, and(6)the growing needs for effective course-management systems and
web services. Without exception, Cardiff University is one of the universities facing the impact too. The Learning
Technology Team, Information Services in the university received increasingly requirements for supporting
virtual classroom and collaboration software (VCCS) in 2012. This analysis therefore collated and clarified the
user requirements for the VCCS in learning and teaching, reviewed VCCS in the market and further
recommended good tools that meet the user requirements. To avoid misunderstanding of the topic, we provided a
definition of Virtual Classroom:
A virtual classroom is an online service to support both synchronous and asynchronous e-learning
activities. It may involve overlapping scopes of interaction such as virtual courses, online meeting,
and online presentation. It supports activities and interactions through functionalities such as video,
voice, chat, screen sharing, webinar, whiteboard, threaded discussion boards and so on.
This article presents the conduction of this study and its findings. It intends to contribute to the
knowledge of understanding user needs for the VCCS.
1
Elluminate Live had been purchased by Blackboard Inc. in 2010 and is named as Blackboard Collaborate.
nd
I.
METHODS AND DATA COLLECTION
This analysis intended to gather opinion-related qualitative information rather than look to measure data
quantitatively. Considering that a VCCS provides many features and it is time consuming to ask participant’s
opinion about each feature, an online questionnaire was designed to collect user requirements for functionalities
of the VCCS. The questions about features of a VCCS were referred to two types of resources. One was the
evaluation documents from the known application websites, such as Blackboard Collaboration generic vendor
evaluation worksheet (e.g., [4]), Adobe Connect product comparison whitepapers [5] and TeamViewer
benchmark ([6]). The other was the academic papers and research studies about the VCCS (e.g., [7], [8], [9]). The
questionnaire covered all features relating to online educational activities. Table I shows an example of the
designed questions for collecting user requirements. Figure I shows an example of questions in the survey. There
were 14 categories in the “Teaching and Learning Feature” theme and 12 categories in the “Administration”
theme.
TABLE I.
Themes
AN EXAMPLE OF FEATURES IN VCCS
Categories
Features
Live Video & Audio
• Test your audio locally
Streaming
• Adjust individual user volume
• Audio/webcam setup wizard
• Drag & Drop reordering
• Play multimedia file (video, audio, Flash, QuickTime, mpg,
Teaching
m4v, etc.)
and
• Play slides
Learning Features
• Resize videos
• Full Screen Video
Capture/
record/
playback
• Moderators can control of sessions can be recorded, stored,
exported or printed
• Moderators can determine if a session/content can be
accessed/ played back
Remote support
• Stand phone lines
• VoIP
• Installation required
Administration
• Remote control
• Support SSL Encryption
Security
• Server software can be hosted securely behind a firewall
Following the online questionnaire, the participant was invited to a one-to-one semi-structured interview
(40-60 minutes) to discuss more details of their requirements according to their responses to the questions in the
survey. The interviewing questions were open-ended and aimed to collect background and extensive information
related to particular requirements from the participant. It focused on the aspects as follows:
nd
•
To understand the participant’s purposes of using the VCCS;
•
To know what teaching and learning activities are involved in the use of the VCCS;
•
To know who are the target users of the VCCS;
•
To know what are essential features of the VCCS for the participant;
•
To know how urgent the requirement is for the VCCS is;
•
To understand more specific subject-related requirements for the VCCS;
•
To clarify any doubts about the questions in the survey;
•
To give the participant an opportunity to interpret their requirements in detail.
Figure 1. An example of survey questions about user requirements for the VCCS features
Twenty-three faculties were invited to participate in this investigation by a recruiting email sending to
domain administrators of the university e-learning system users. Fifteen volunteers from eleven faculties have
participated in the study. Six out of fifteen participants have experience of using VCCS, and nine participants
stated that they are planning to use a VCCS tool in learning and teaching activities.
The collection of available VCCS products in the market was based on four groups: (1) tools that we
have already known, (2) tools recommended by e-learning professional interest groups and organizations (e.g.,
JISC funded projects [10]), (3) tools recommended in research papers and academic journal articles in education
settings (e.g., [11], [12]), and (4) reliable awards for learning technology products (e.g., [13]). It included 46
VCCS tools on the list. Based on the availability, the VCCS products were categorized into three groups: 16 free
register version tools, 17 free trial version tools and 9 commercial version tools. The nine commercial version
tools were not tested because of one of the reasons: (1) did not provide a trial version, (2) requested to register
credit card information to try a trial version, (3) requested certain membership.
The findings were based on evaluating tools in the group of free register version software and the group
of free trial version software. The evaluation of the VCCS products is not comprehensive on the following
aspects considering limitations of the availability of testing facilities, testing members and the investigation
duration:
nd
•
Using mobile devices to access the VCCS was not tested;
•
The VCCS running on Windows XP SP3 and Mac OS X 10.6.8 was tested;
•
The VCCS running on web browsers (Firefox 13.0.1, Firefox 12, Safari 5.1.7, Google Chrome 20, IE
8) was tested;
•
A large group of attendees (over 10) synchronously joining in a live session was not tested.
•
It did not test every feature of a VCCS product but focused on testing the features required by
participants;
•
Some
VCCS
applications
(e.g.,
Voluxionhttp://www.voluxion.com/;
Zoho
Meeting
https://www.zoho.com/meeting/) were dismissed because they either did not provide a free trial version
or needed credit card information to register. It does not say this type of applications is not good.
II. FINDINGS
A. User requirements for the VCCS
The data showed that the participants have a positive view about the use of VCCS. Six key aspects of
requirements relating to the use of the VCCS were emerged: (1) supporting online training and tutoring, (2)
supporting online lecture presentation, (3) supporting virtual meetings, (4) enhancing flexible learning
opportunities, (5) providing flexible teaching approaches, and (6) supporting subjects and special needs learners
related e-learning. The core theory occurred here is Flexible Learning. According to Collis and Moonen [14], the
theory is reflected on three dimensions from the data. At first, users required for flexible access learning content.
Some participants mentioned that more and more students (including campus-based students) asked whether the
teaching sessions could be available online for them to access later. Secondly, users expected flexible learning
style regarding to time and location. Participants stated that distance learners who take online courses need to join
a live session or access to various online resources from different locations and time zones at the same time.
Hence, as the third point, schools want e-learning systems to support online learning activities, and teaching staff
want more flexible approaches to delivering e-learning courses. Without doubt, the requirements for the VCCS
are learner-centered. People want a flexible learning environment cross the traditional classroom so that teaching
and learning can fit into their lifestyle.
Figure 2 presents an overview of user requirements for the VCCS. It shows the main uses of the VCCS
in schools/departments, the priority of the requirements in each use, and the fundamental features in relation to a
use of the VCCS.
nd
Figure 2. An overview of user requirements for the VCCS
There are three distinct requirements particularly in relation to learning and teaching activities. The first
is supporting online training and tutoring. This is the most urgent requirement. This group of users was from
libraries, the lifelong learning centre, and schools that run e-learning courses. The requirements represent on (1)
having an easier approach for learners to join in a training session no matter their locations, (2) a live training
session can be recorded by moderators, (3) recorded training sessions are accessible for learners, (4) the presenter
is able to share screen and application, and (5) easily organizing group work or group discussions in a live
session. This type of requirements was categorized according to common required factors as follows:
•
Small numbers of attendees (maximum 20 attendees);
•
Learners who are unable to attend a session on the site and/or at a specific date/time can join in a
session no matter their locations and their time (e.g., distance learning students, business or the
individual who have limited time);
•
Learners may be at different levels (e.g., different language background in a language training course,
different levels of literacy skills in a library databases searching course);
•
1-2 moderators may lead a session;
•
A variety of learning activities to engage learners (e.g., play presentation slides, quizzes, access to
external web resources, play a video);
•
Supporting group work and discussions;
•
One-to-one, one-to-many or many-to-many interactions (e.g., allow attendees to chat to each other).
Table 2 displays an example of the essential features in the VCCS if it is used for the purpose of
supporting small group online training or tutoring (NB: the listed information is not integral). In the table, the first
column refers to the key requirements stated by participants; the second column lists the functionalities in a
VCCS tool; and the third column shows the category that we used to design the questionnaire (see Table I, the
nd
second column). The VCCS features such as supporting interaction and a variety of ways to engage users,
recording, chat, video/audio, screen/application sharing are crucial.
TABLE II.
AN EXAMPLE OF USER KEY REQUIREMENTS FOR THE VCCS IN THE USE OF SUPPORTING ONLINE TRAINING/TUTORING
Key requirements
Required VCCS features
Feature categories
Be able to schedule
Upload files (e.g., presentation slides, quizzes,
File transfer
sessions
documents, multimedia files)
Email attendees and guests to a scheduled
Communication
session
Allow group work
Breakout rooms
Breakout rooms
Whiteboard
Whiteboard
Textual public messaging
Communication
Textual private messaging
Audio/video
Be able to play
File transfer
File transfer
Support various types of multimedia files
Live
multimedia files
Be able to see/show
video
&
Audio
streaming
Screen sharing or Remote control
Application/Screen sharing
Sessions recording
Capture/record/playback
a attendee’s screen
Be able to record
the training session
Secondly, supporting online lecture presentations is a fairly urgent requirement. This group of users need
the VCCS features such as supporting a large number of attendees connected synchronously, recording, playing
high-quality presentations. It is related to the user who encountered issues such as it is difficult to check students’
attendance, the shortage of big lecture theatres in the university, and the lack of flexible learning opportunities for
students. This group of users is from schools that run e-learning courses and that deliver lectures to a large group
of students. In particular, the schools run e-learning courses to distance students need a better VCCS solution
urgently. This type of requirements was categorized based on the factors as follows:
•
Large numbers of attendees (minimum 50 attendees, normally over 100 attendees);
•
Live video presentations to geographically dispersed learners;
•
High quality presentations;
•
Record live sessions and edit it before publishing for on-demand access;
•
The lecturer/presenter leads a session and moderators moderate in due time;
•
Using learning activities to engage learners during the session (e.g., quizzes, polling, survey);
•
One-to-many or a few-to-many interactions (in a lecture, it has little time for group discussions).
Thirdly, supporting virtual meetings was asked by participants too. The users expected the VCCS allows
anyone, anywhere to share their desktop and ideas with others instantly. This use is not as urgent as the above two
nd
required uses. Features such as supporting file transferring, recording, host meetings, chat, video/audio, screen
sharing are important. This type of requirements was categorized based on the factors as follows:
•
Attendees are at different locations and/or time zones;
•
Attendees communicate through audio;
•
Multiparty video (e.g., attendees can see each other);
•
Play presentation slides;
•
Whiteboard and/or screen/application sharing;
•
Many-to-many style of interaction.
In addition to the above, there are three fundamental requirements at the pedagogical level. The three
requirements are intrinsically linked in relation to the main uses of the VCCS. They are equally important in a use
of the VCCS. In a specific learning and teaching activity, participants expected the use of the VCCS to “enhance
the students learning opportunities”, for example,
•
The features allow students to attend a session from different locations and/or time zones,
•
Allow students customise the screen size/font/colour,
•
Online chat,
•
Communicate through video/audio,
•
Use whiteboard,
•
Play quizzes, and
•
Allow users to access from mobile devices.
They also expected that the VCCS provide features to support flexible teaching approaches, such as
•
allowing educators to invite external people into a session,
•
having different levels of controls,
•
easily managing online sessions,
•
tracking and assessing students’ learning outcomes, and
•
easily working with an existing virtual learning system.
Regarding to a specific subject or a specific need learner’s requirement, users expected the VCCS is able
to support an extensible architecture that allows the addition of new tools to meet these specific needs, for
example, allowing users to key in complex mathematics symbols or allowing learners with dyslexia to listen to an
audio recording of a session.
B. Reviewing of the VCCS
After reviewing thirty-three available tools in market, it found that three free VCCS applications: Any
Meeting (http://www.anymeeting.com/), Team Viewer (http://www.teamviewer.com), Wiggio
nd
(http://wiggio.com/) and five commercial applications: Adobe Connect
(http://www.adobe.com/uk/products/adobeconnect.html), Blackboard Collaborate
(http://www.blackboard.com/platforms/collaborate/overview.aspx), Microsoft Lync
(http://lync.microsoft.com/en-gb/launch/Pages/launch.aspx), om Novia (http://www.omnovia.com/), WizIQ
(http://www.wiziq.com/) meet most user requirements. In general, they are easy to use, run on all supported
operating systems and web browsers, and provide reliable service and flexible features.
Considering to support online training and tutoring, all eight applications have “the most needed”
features, although each lacks of some features. For example,
•
Any Meeting does not allow attendees to upload/download files and does not have breakout rooms.
•
Team Viewer does not have breakout rooms, polling, quizzes, and action icons (e.g., hands-up) to
support interaction.
•
Wiggio has a problem of recording/playing back in the current free version.
•
omNovia does not support Mac users.
•
WizIQ does not have quizzes and application sharing.
•
Adobe Connect is unable to export polling results and not allow attendees to resize video.
•
Microsoft Lync does not support breakout rooms.
•
Blackboard Collaborate does not allow attendees to customise font colour, but its features as a whole
are slightly better than those features in Microsoft Lync.
In relation to supporting online lecture presentations, Any Meeting is the only free application that
supports a large group of attendees (up to 200 attendees). Both Blackboard Collaborate and Microsoft Lync
provide many required features. Blackboard Collaborate supports additional tools (e.g., complex mathematics
symbols) and more types of multimedia files than Microsoft Lync.
In relation to supporting virtual meetings, as a free tool, Team Viewer provides all key features the users
required. Wiggio is good at supporting work in groups and collaboration. For the commercial tools, Adobe
Connect, Blackboard Collaborate and Microsoft Lync all satisfy the user current requirements.
In accordance with pedagogical support, to provide flexible learning opportunities and teaching
approaches, Any Meeting is better than Team Viewer and Wiggio on the aspects of supporting web tour, polling,
recording and tracking students’ learning outcomes. However, if the user is trying to use iPad/iPhone, Team
Viewer as a whole service provides more useful features than Wiggio. OmNovia and WizIQ are not as good as
the other three commercial applications. Adobe Connect is not as good as Blackboard Collaborate because it
lacks of some flexible features such as allowing attendees to customize chat font type and video size, and
allowing the user to export polling results.
Reviewing the application as a complete service, both Blackboard Collaborate and Microsoft Lync
provide many flexible and useful features. However, Microsoft Lync is not as good as Blackboard Collaborate in
its component tools such as supporting breakout rooms, supporting tracking reports and assessing students’
nd
learning outcomes, supporting extensible tools for specific needs learner and subjects, supporting multimedia file
types, supporting multiple language interfaces, and providing a solution of integrating with the existing Learning
Management System (LMS). Compared to Blackboard Collaborate, Microsoft Lync is a proper option for general
collaborative requirements within an organization, but is not the best software option for supporting learning and
teaching activities specifically. Figure 3 displays the recommended VCCS applications regarding to user
requirements.
Figure 3. An overview of good VCCS tools and user requirements
III.
CONCLUSIONS
It is apparent that no single application meets all user requirements. In accordance with different subjects
and learning activities, there are different aspects and degrees of requirements for the VCCS features. One
application may be more suitable than the other in relation to one or two particular requirements for the VCCS in
the school. Based on the analysis, recommendations provided for users include:
•
The users need to be aware that each tool has some weakness. When choose a VCCS application, they
need to clear the most essential features in the VCCS in relation to their learning activities.
•
In the short term, the schools that urgently need the VCCS (such as supporting tutorials or online
trainings for a small group of distance students) should adopt a free application such as AnyMeeting
until a centrally supported solution can be put in place.
•
For the users who want to use the VCCS to run virtual meetings or support communication and
collaboration, TeamViewer is a good choice and Wiggio satisfies the users who particularly need the
VCCS for supporting work in groups.
•
For the schools that run e-learning courses for distance students (especially a large group of learners) or
want to apply the VCCS to support lecture presentations, they may use AnyMeeting to support learning
and teaching activities before a centrally supported solution is available.
nd
•
Where there is a specific requirement to support specific subjects or specific needs of learners and
providing a flexible learning environment, Blackboard Collaborate clearly meets more user
requirements than Microsoft Lync.
•
Schools that have prepared funding for purchasing a VCCS tool should consider Blackboard
Collaborate.
•
Should any schools use Microsoft Lync or Blackboard Collaborate in the short term then their
experience should be closely monitored to provide a comparison between the two applications.
Overall, it is hoped that this article helps people better understand the user requirements for the VCCS
and further provides a sound in the support strategic decisions where it pertains to these technologies.
ACKNOWLEDGMENT
I would like to express my special gratitude to the contribution of participants.
REFERENCES
[1]
S.R. Hiltz and M. Turoff, “Video plus virtual classroom for distance education: experience with graduate
courses”. The Distance Education in DoD Invitational Research Conference, National Defense University,
Washington DC, USA, 11-12 February 1993. Invited paper.
[2]
J.L. McBrien, P. Jones, and R. Cheng , “Virtual spaces: employing a synchronous online classroom to
facilitate student engagement in online learning”. International Review of Research in Open and Distance
Learning, vol. 10, no. 3. 2009.
[3]
S. Howell, P. Williams, and N. Lindsay, ‘Thirty-two trends affecting distance education: an informed
foundation for strategic planning”. The Online Journal of Distance Learning Administration, vol. VI, no.
III, Fall 2003, pp. 23-26.
[4]
“Virtual classroom evaluation 2011”. IT Services at UQ, The University of Queensland, Australia, 2011,
Web. Accessed 15 July 2012.http://www.its.uq.edu.au/index.html?page=169750&pid=144707
[5]
“Competitive comparisons”. GetConnect, Adobe Connect partner, 2011. Web. Accessed 3 April 2012.
http://www.getconnect.com/resources/competitiveprodcompare/
[6]
“Evaluation-Benchmark”.
TeamViewer,
[n.d.].
Web.
Accessed
3
April
2012.
http://www.teamviewer.com/en/res/excel/TeamViewer-Benchmark-en.xls
[7]
F.C. Serce and S. Yildirim, “A web-based synchronous collaborative review tool: a case study of an on-line
graduate course”. Educational Technology & Society, vol. 9, no. 2, 2006, pp. 166-177.
[8]
R.E. Griffin, D.M. Friedman, and J.S. Miko, “Discover how virtual classroom software (VCS) can
revolutionize your distance education classroom”. Proceedings of the EDUCAUSE Mid-Atlantic Regional
Conference, Baltimore, Maryland , 17-19 January 2007.
[9]
S. Sharma and S. Gupta, “The virtual classroom: the role of ICT in open and distance learning”.
International Journal of Computer, Communication and Emerging Technology (IJCCET), vol. 1, no. 1, pp.
8-12, April 2012.
nd
[10]
P. Chatterton, “Designing for participant engagement with Blackboard Collaborate: a good practice guide to
using Blackboard Collaborate to support teaching, learning & assessment, co-operative working and
conferences”. The Joint Integrated Information Environment Committee (JISC). Web. Accessed 19 May
2012. http://www.jisc.ac.uk/whatwedo/programmes/elearning/collaborateguidance.aspx
[11]
K. Michael, “Virtual classroom: reflections of online learning”. Campus-Wide Information Systems, vol.
29, no. 3, 2012, pp. 156-165.
[12]
P. Jones, M. Kolloff, and F. Kolloff, “Using web conferencing to promote learning in online courses:
resources, tools, and methods,” in Proceedings of Society for Information Technology & Teacher Education
International Conference 2012, 5 March 2012, Austin, Texas, USA, P. Resta, Ed., Chesapeake, VA: AACE,
2012, pp. 481-486.
[13]
“Global e-Learning Awards”, Learn Tech Conference & Exhibition, Web. Accessed 26 July 2012.
http://worldeducationcongress.com/learntech/Awards_Categories.html
[14]
B. Collis and J. Moonen, Flexible Learning in a Digital World, 2nd ed., London: Routledge and Falmer,
2004.
nd
The use of learning technologies to facilitate engagement in an online course
Dilani S. P. Gedera
TEMS Education Research Centre
The University of Waikato, New Zealand
[email protected]
P. John Williams
[email protected]
Abstract:
E-learning is becoming increasingly popular in many countries for its flexibility in terms of time, place
and pace. Research affirms that learning technologies support interaction and collaboration among learners
and improve learning outcomes. However, current practices of e-learning are not without constraints and there
is a need for empirical research to assist practitioners in determining the best uses of learning technologies.
This paper seeks to develop an understanding of students’ experiences and their perspectives of learning with
the educational technologies of ‘Adobe virtual classroom’ and ‘Moodle’ that facilitated activities in a university
course. The study was conducted using a case study method over a period of one semester. With Activity Theory
as its research framework, the research methods of this study include individual interviews, online observation
and document analysis. This paper includes some of the initial findings of the research and a brief discussion on
how the educational technologies facilitated students’ engagement in this course. This may inform practitioners
of the pragmatic constraints and affordances of existing technologies, learning activities and strategies used in
online learning environments.
Keywords-component; learning technologie; engagement; e-learning; affordances; constraints; Activity
Theory
I. INTRODUCTION
E-learning is becoming increasingly popular in many countries for its flexibility in terms of time, place
and pace. Research affirms that learning technologies support interaction and collaboration among learners and
improve learning outcomes. Studies that examine how technology can be used in educational contexts have
focused on potential benefits of learning technologies in relation to teaching and learning. For instance, Holmes
and Gardner [2] assert that e-learning offers unique opportunities for teachers and learners to enhance their
teaching and learning experiences via virtual environments that help not only in delivery but also in the
application of knowledge.
In addition to the pedagogical potential of e-learning, some researchers have focused on the changes
that are taking place with the emergence of e-learning. The shift from teacher-centred learning to studentcentred learning in various contexts [1][7] and also the way students learn and interact in learning environments
[6] are some of these changes. For instance, access to information regardless of time and place enables learners
nd
to ‘explore education’ and communicate with peers and the outside world through various media ranging from
print to video [6].
In the context of New Zealand, The Ministry of Education stresses the importance of e-learning in
education. According to NZCER [7] “E-learning can improve understanding and encourage deeper learning, if
there is careful course design and choice of technology in relation to learning objectives that aim to encourage
deeper learning”. As suggested by the e-learning Advisory Group [3], technology does not offer a complete
solution for a transformative education; rather the practitioners should focus on the potentials of learning
technologies that can enhance students’ learning experience and performance as well as the constraints of these
technologies that inhibit their performance. These affordances and constraints of technologies in education
should be thoroughly considered for a successful implementation of e-learning.
This paper seeks to develop an understanding of students’ experiences and their perspectives of
learning with the technologies –Adobe virtual classroom and Moodle that facilitated activities, and how these
technologies facilitated students’ engagement in a fully online university course. This study defines learner
engagement as students’ active participation in e-learning activities such as discussion forums and virtual
classroom activity in achieving learning goals in an online environment.
II. RESEARCH CONTEXT AND RESEARCH QUESTIONS
This paper is based on a case study focused on a fully online course that was offered in the first
semester of 2012 academic year. A total of seven participants that comprised six students and their lecturer
participated in this study. The research questions that guided the data collection and analysis of this research
were:
•
What were students’ perspectives on their experiences of learning with Adobe virtual classroom and
Moodle learning technologies that facilitated activities in a fully online university paper?
•
How did the educational technologies affect students’ participation in e-learning activities?
III. RESEARCH FRAMEWORK, METHODS AND DATA ANALYSIS
This study was carried out as a case study over a period of one semester and the methods of data
collection included individual interviews, observation of online activities and document analysis. The online
learning activity was the focus of data gathering and Engeström’s [10] Activity Theory was used as the research
framework.
Activity Theory is derived from socio-cultural and socio-historical theories and can be considered as a
philosophical framework that is used to study practices and processes of human beings [4]. Particularly,
Engeström’s activity systems analysis enables researchers to observe the interactions that take place among
individuals and the environment and how each affects one another [5]. The elements of an activity system
include subject, object and other mediators such as tools, rules, community and division of labour.
In capturing participants’ views on how these educational technologies facilitated students’
engagement, particular attention was given to the tool mediation principle of Activity Theory- that is human
activity is mediated by several tools [9]. These tools can be physical—a computer or a book, conceptual– a
mental model, a plan or a strategy, abstract— a language, or even a virtual tool— functions of a website.
nd
Mediation of tools plays an important role in shaping how human beings act and interact with the world [8].
Focusing on the tool mediation tenet of Activity Theory, the sub-themes that emerged under the element Tools
were mainly considered for the analysis. In the case of this paper, tool mediation refers to the use of Adobe
virtual classroom that facilitated a synchronous activity in which students presented their research to the
members of the class, and the Moodle learning management system that facilitated an asynchronous online
forum activity in this course.
In coding data, relevant units from transcribed interview texts, observational notes and documents (i.e.
course outline) were identified and categorized according to the elements of Activity theory as a method of
typology. In categorizing these codes according to the elements of Activity Theory, Nvivo was used as a data
management tool.
IV. FINDINGS AND DISCUSSION
The activities observed and used in the analysis of data included Adobe virtual class that facilitated a
synchronous activity and Moodle that facilitated asynchronous forum discussions as well as the design and the
delivery of the contents of the course. Thus, the findings that are related to these educational technologies and
other tools that mediated students’ engagement in activities in this course are illustrated in the following
sections.
A. The Adobe virtual classroom
The Adobe Connect Pro virtual classroom that was used as the tool in facilitating the collaborative
synchronous activity in this class was a web-based technology. This activity was carried out as an individual
assignment and represented 30% of the student assessments. In order for students to access it, a unique URL and
a password was provided by the lecturer.
In placing the Adobe virtual classroom activity in Activity Theory framework, the subject represents
the student(s) who are the focus of this study. The object is the purpose of an activity which can be a motive or a
problem space. In this case, the students’ purpose was to present their research to an audience (other members of
the class). The mediating tools that were used in this activity in order to transform students’ object include
physical tools- computers, mental tools- learning strategies, models and virtual tools- functions that were
available on Adobe virtual classroom. The rules for this activity included the duration of the presentation (10
minutes), relevant literature and references (following APA format) and a written script or notes (1500 words).
The community of this activity includes the facilitator who is also the lecturer of this course and the members of
the class. Division of labourdefines the students’ responsibilities. As part of their responsibilities, one peer had
to review the allocated student’s presentation and the notes before the actual activity and also the peer was to
raise three questions to be discussed after the presentation.
The findings suggest that the students’ experiences of learning with Adobe virtual classroom were
associated mainly with the affordances and constraints of this learning technology. As pointed out by the
participants, the affordances of the Adobe virtual class allowed them to interact in real time. Denoting the value
of synchronous tools that enable them to see and hear each other in real time and the immediate responses they
receive when they engage in collaborative activities, Alex suggested that:
nd
I would like to see more synchronous. Only because I like seeing people when I’m talking to them and stuff like
that. I like that backwards and forwards that can happen very easily in that environment (Alex, interview 2)
Students also pointed out that the visual images are important in communication and without them,
they probably find it quite hard to feel like they belonged to a learning community.
Another feature of virtual classroom that supported students’ engagement was the ability to have an
oral discussion in real time right after each presentation. It was observed that having a discussion after each
presentation allowed the students to clarify the issues related to the topic immediately, as well as provide some
instant peer feedback. When Alex was asked what he thought about the reviewing of notes and facilitating a
discussion after each presentation, he stated that “I think it caused us slightly deeper interaction with what the
others have done. I thought that was quite useful”(Alex, interview 2). As a group, students could also support
each other by giving words of encouragement after their presentations, for example “very interesting”, “welldone”and “excellent presentation”. These video and audio features facilitated two way communications among
the students and thereby created a sense of belonging to a learning community.
Apart from the video and audio functions, Adobe virtual class also allowed the students to have a textbased chat during this activity. This was particularly useful when they had questions to ask from a particular
person in private or in public as well as to have a chat before the facilitator (lecturer) joined the group. An
example of a text-based chat is shown below.
Debbie: Why am I coming up as guest?
Debbie: you must have heard me
Alex: Hi Debbie - you've come in as Guest, if you close out and come back in you should get the chance to put in
your name. I'm backing you, so I hope your nervousness isn't too bad
Alex: All the different styles make it interesting :) imagine if they were all the same :(
Alex: Hi Debbie - not sure why that didn't work... I'll investigate
Debbie: I’ve just tried coming in again but I don’t see a guest tag now
Alex: It has remembered you and brought you back in as Guest again - maybe Garry can tell you how to fix that
Richard: Debbie - quit your browser and re-enter with your name and not guest
Alex: Hi Debbie - did you hear that
Debbie: Yes
(7 May observation)
On the other hand, students also felt that there were limitations with the Adobe virtual classroom.
Students mentioned that they were attracted to this course for its flexibility, as it was fully online and enabled
them to have flexibility in terms of time, place and pace. However, as Richard, the lecturer pointed out, the
inclusion of synchronous activities like the research presentations using Adobe virtual classroom needs careful
planning in terms of time, as some of the students come from different countries and when the time zones are
different, it is hard to coordinate synchronous activities. On that particular day and at that particular time if the
nd
internet connection is not stable, the students may not be able to participate in the activity. This was evident in
the case of Gail who participated from a Middle Eastern country. Due to slow speed internet connection that was
caused by an unstable political situation in the country, Gail couldn’t hear what the others were saying and also
she couldn’t do her presentation or facilitate and join discussions. Although Richard gave her a one-to-one
session to present her research via Skype the next day, it was a disappointing and frustrating experience for her.
Richard described that:
I had her notes and I had her PowerPoint slides and I also had a Skype conversation with it. The issue was
there, the bombs are going outside the window and probably it had something to do with it, but from her point
of view, it was frustrating because she had prepared and she did a good job (Richard, interview 2).
This suggests that although online courses provide more flexibility to learners and the ability to interact
with each other in real time, synchronous activities like Adobe virtual classroom need careful planning in its
implementation especially with back up plans, as unexpected situations like weather conditions and technical
issues can hinder the execution of activities.
The findings indicated another significant limitation of this tool is that if the number of students is
high, it gets harder to allocate time slots. As Richard said, “the more students you have the harder it is. Also
how many times you have to do it” (interview 2). Considering the constraints of Adobe virtual classroom
Richard appeared to be unsure of its best use and he mentioned that “in using Virtual classroom as I’m using it
more ….as I go along…….I am still little bit unsure about its best use (Richard, interview 1). Participants’ views
also suggested that having limited capacity for only one speaker to talk at a time, discussions take a longer time
and also the participants may have to repeat their utterances many times if two people talk at the same time.
However, it was interesting to find out that students still preferred more synchronous activities despite
its limitations because the students could have more “human interactions” where they could use “a lot of cues”
(Alex, interview 2) like facial expressions as well, as it allowed the participants to interact with each other in
real time that enabled them to have a two way communication. In Eddy’s point of view “if there was a little
more any form of synchronous…it may have helped get to know one another earlier on (Eddy, interview 2).
Alex pointed out that although the virtual classroom “wasn’t flowing more like we can talk like we would be
sitting in front of each other just in person….but when face to face is not possible, it’s better than not having it.
(Alex, interview 2).
B. Moodle
This fully online course was delivered via the university learning management system—Moodle.
Therefore, in designing this course, the lecturer used several strategies that were facilitated by Moodle. The
findings suggest that these strategies as conceptual tools facilitated students’ engagement in this course.
As explained by the lecturer it is vital for online courses to be well structured with a similar pattern in
each week or module. The uniformity of the course allows the learners to have a positive experience of learning
with learning technologies, as the students did not have to struggle to find materials and resources in separate
folders. As one of the strategies, all the resources that included reading materials, YouTube clips, helpful tips
and guidelines for writing and assignments were embedded in the texts or hyperlinks on the course Moodle
page. As Richard said:
nd
There are no boxes or folders full of papers for students to work their way through, but all are embedded within
the texts or hyperlinks and everything follows a logical progression (interview 1)
This strategy seemed to make students’ access to information as easy as possible, as they didn’t “have
to wrestle with the interface or finding resources to be able to learn”(Richard, interview 1). The importance of a
structured course was also highlighted by the students and they appreciated that the lecturer “has been a very
good coordinator and his work is structured”(Christine). Students also described that all the information is there
and they can read in their own time.
Another strategy the lecturer used in the design of this course was creating several spaces for students
to interact and communicate. These spaces comprised class news and notices, private and public communication
spaces, sharing spaces, peer support spaces and Q and A spaces for each module. By creating several spaces for
communication, the students were given a choice to suit their needs. In the lecturer’s point of view, these
strategies should be deliberately implemented and one strategy that suits one group may not work for another;
therefore, it is important to consider the needs of that particular learning group in designing online courses. As
pointed out by a student, these spaces were helpful when they needed help with their assignments. In situations
where they needed assistance, more capable peers came in and guided them through the process.
I sent out a question on the public Question place. Alex came in and sent me an email how to do it. To me, it was
little bit like a miracle because I had no knowledge of how to do either of those the Movie maker or the
YouTube. And without instructions that Alex emailed, I couldn’t have done it (Christine, interview 2)
In this case, Christine was struggling with one assignment where she had to create a 3-4 minute
presentation to be uploaded to YouTube (or equivalent online multimedia tool). Alex, having the technical
knowledge, offered help. Other general questions that were shared and answered on the public question space
included questions on internet speed requirements, editing functions of Moodle discussions, assignment due
dates, reference styles and computer brands.
well used space in this course was the sharing space. This space was continuously filled with
information about software—PDF reader, screen capture, as well as information on embedding audio into
postings, using shortcuts, uploading URLs and docking blocks. Students acknowledged that sharing of ideas
among peers about technology that can help them in their teaching and learning is particularly useful in their
jobs although they are not tech savvy. Debbie stated:
These are wonderful little instruments that everybody else knows and I don’t…so we share things that might be
helpful and it’s a great advantage. I’m never going to be brilliant at technology, but I might be able to make it
useful to me in my job little bit more I think (interview 2)
The participants’ views suggested that the sharing of information and ideas helped them to get
connected to each other as a learning community and increase their engagement with learning. Overall, the
strategies that were used in the design of this course, as a conceptual tool, allowed the students to interact and
communicate with each other in public or in private, choosing the method that suits them best. The question and
answer space was in use to get help when students needed assistance with their assignments. The sharing space
helped them to interact with each other while providing information on useful instruments that can be used in
teaching and learning. As a conceptual tool, these structural strategies created by the lecturer acted as a mediator
nd
in making close connections among students allowing them to have a sense of belonging to a learning
community.
In terms of the asynchronous forum discussion activity that was facilitated by Moodle, student
participants reported both positive and negative opinions. One common idea shared by the students was that
forum discussions as an asynchronous activity gives more time for them to reflect especially if students are shy
or have problems with English as a second language. In Alex’s opinion, those second language speakers, when
they are put on the spot it’s a challenge, but on forums they? have time to reflect and be prepared (Alex,
interview 1). Debbie described that as a shy person she never talked in face-to-face tutorials during her first
degree, as she thought whatever she says has already been said by other students and it’s not going to add any
value. She believed that in forum discussions everybody says something , and online forums are an excellent
method for discussions.
However, students felt that the forum discussions in this course were rather “official” and it is slow to
get to know people, as they do not get to see people in forum discussions. As explained by Richard, humour
plays an important role in online discussions. He explained:
Because I’ll be nervous…..as they are...you’ve got to make them see you as a person who they can talk with
…and identify the person with it. Humour is part of that...and as you go along…they know when you pull their
leg and you know...that’s really important…so they feel comfortable and relaxed to be able to interact with me
and others. And responding to those every day (Richard, interview 1)
He believes that humour helps students to lessen nervousness and increase interactions among students.
However, Alex emphasized that in forum discussions, it is hard to have a sense of humour without physical
cues.
I’m thinking of a couple of discussions we’ve had where I haven’t quite been sure what perspective the person
was coming from whereas you see them saying it, you can often judge a bit more from some of the physical cues.
I mean when I write, I write in some ways with the tone I have in my mind, but it can be sometimes sarcastic.
You can make it come out, but that doesn’t fully come out with the intonation and the timing and everything
(Alex, interview 2)
As a learning experience, students did not fully enjoy the discussions, as they thought that physical
cues were missing. Moodle as a virtual tool allowed the lecturer to embed resources within texts as well as
design a well-structured Moodle page for this course.
C. Language
In this case, Fiona who is a second language English speaker seemed to have issues, as at times she
could not understand what the others were talking about.
I think sometimes I feel that I’m lost in discussions. Actually in this paper, I don’t participate much because as I
said I’m lost in this conversation. I need time (interview 1)
Since Fiona could not follow them she mentioned that she only read the discussions. When she was
asked for the reasons why she couldn’t follow them, she said that, “When they used slang I didn’t understand. I
try to understand the general meaning. Sometimes they use short forms like letters. Sometimes even Richard
nd
uses this” (interview 2). She also had difficulties in doing assignments, as she found it hard to understand what
exactly needed to be done, but she managed to clarify things with the help of her fellow classmates. She
mentioned that “I found difficulties in doing assignments because I didn’t understand, but I asked the others and
the things became clearer” (interview 2).
When Fiona was participating in Adobe virtual classroom activity, she had some difficulties with
grammar and pronunciation (7 May observation). She seemed to be reading from her slides and she mentioned
that she was wondering whether her peers could understand her presentation. As part of the Adobe virtual
classroom activity, students reviewed one of their peer’s presentation notes and facilitated a discussion by
asking three questions from the presenter right after their presentation. Providing her views regarding this
process and how language became an issue in this case, Debbie described that:
Well from my point of view the person who asked me questions, that was Fiona. Fiona has issues with English
as a second language and the questions she asked me were not exactly around my presentation. They were more
about who I was and what I was doing, so I didn’t find her questions particularly useful (interview 2)
The findings suggested that Fiona appeared to be struggling with English as a second language in this
course. In terms of tool mediation, it is apparent that language as an abstract tool mediated Fiona’s engagement
in both forum discussions and Adobe virtual Classroom activities. However, in her view since the course was
fully online she had an advantage, as she had more time to think and reflect. Fiona compared her online learning
experience with a face-to-face class and emphasized that “because of language, it is better to learn online
because I have my time to read” (Fiona, interview 1).
V. CONCLUSION
This paper highlighted students’ experiences and their perspectives of learning with technologies –
Adobe virtual classroom and Moodle that facilitated activities, and their influence on students’ engagement in a
fully online university course. In this context, the affordances of the synchronous Adobe virtual classroom, as a
virtual tool, facilitated collaboration and helped students in creating a sense of belonging to a learning
community particularly with the video and audio features. However, the data also suggested that when
interacting in real time some practical issues such as time zones need to be considered in using synchronous
learning technologies. In terms of the asynchronous forum discussion activity that was facilitated by Moodle,
participants’ views comprised both positive and negative opinions. Moodle discussions as an asynchronous
activity, gave students more time to reflect. However, in forum discussions it seemed that it takes time to get to
know people and also it is hard to have a sense of humour without physical cues. In terms of the design of the
course, Moodle as a virtual tool allowed the lecturer to create a well-structured course where all the resources,
video files and other documents were embedded in texts and hyperlinks and organized according to a logical
progression that enabled the students to have a positive experience of learning with learning technologies.
Students’ voices indicated overall satisfaction in learning with learning technologies and achieving their
learning goals in this course.
References
[1] B. C. Lee, J. Yoon, and I. Lee, “Learners’ acceptance of e-learning in South Korea: Theories and results,”
Computers and Education vol. 53, pp. 1320-1329, 2009.
nd
[2] B. Holmes and J. Gardner, E-learning: concepts and practice, United Kingdom: SAGE, 2006.
[3] E-learning Advisory Group, Highways and pathways: Exploring New Zealand's e-learning opportunities,
Wellington: Ministry of Education, 2002.
[4] K. Kuutti, “Activity theory as a potential framework for human computer interaction research,” In Context
and consciousness: Activity theory and human computer interaction, B. Nardi, Ed. Cambridge: MIT Press,
1995, pp. 17-44.
[5] L. C. Yamagata-Lynch, “Activity systems analysis methods: Understanding complex learning
environments,” New York: Springer, 2010.
[6] M. S. Desai, J. Hart, and T. C. Richards, “E-learning: paradigm shift in education,” Education, vol. 129, no.
2, pp. 327-334, 2008.
[7] New Zealand Council for Education Research, Statistical profile of Maori in tertiary level education and
engagement in e-learning, Wellington: New Zealand, 2004.
[8] V. Kaptelinin, B. Nardi, and C. Macaulay, “The activity checklist: A tool for representing the space of
context,” Interactions, vol.6, no. 4, pp. 27-39, 1999.
[9] V. Kaptelinin,. “Activity theory: Implications for human-computer interaction,” Context and Consciousness:
Activity Theory and human-computer interaction, B. Nardi, Ed. Cambridge, MA: The MIT Press, 1996, pp.
103-116.
[10] Y. Engeström, “Learning by expanding: An activity-theoretical approach to developmental research,”
Helsinki: Orienta-Konsultit, 1987.
nd
Using Activity Theory to understand contradictions in an online university course facilitated
by Moodle
Dilani S. P. Gedera
[email protected]
P. John Williams
[email protected]
Abstract:
Activity Theory can offer insights into learning processes that are facilitated by Learning Management
Systems. Contradictions, as a basic principle of Activity Theory, assist in identifying the tensions and conflicts
that emerge in systems of online learning environments. Using Activity Theory as its research framework, this
study focuses on the contradictions that emerged in the form of tensions, frustrations, misunderstandings and
miscommunication in a fully online university course in New Zealand. The data collection methods of this case
study included individual interviews, online activity observation and documents analysis. Outlining some of the
findings of the study, this paper will discuss how students’ participation in learning activities facilitated by
Moodle was affected by these contradictions.
Keywords-Component ; Moodle; contradictions; Activity Theory; Learning Management Systems
I.
INTRODUCTION
Technological advancements in education have provided students with flexible and collaborative learning
opportunities. For example, Learning Management Systems (LMS) have the ability to facilitate synchronous and
asynchronous activities in which students have flexibility in participating in learning activities in terms of time,
place and place.These systems also enable collaboration and interaction among students.
LMSs are also known as “learning platforms”, “course management systems”, “instructional management
systems” and “distributed learning systems” [1], and some of the common LMSs are Moodle, Blackboard,
Oncourse, Angel andSakai. These systems are used to enhance face-to-face university courses as well as fully
online courses. LMSs generally provide various tools for course administration and delivery and some of these
functions include synchronous and asynchronous communication (chat and discussion forums), development
and delivery of content (links to the internet resources), assessment (submission of assignments, quizzes), and
management of students and class (enrolment, class list) [1]. Although LMSs are considered a powerful
technology that can handle a range of aspects of learning processes with its functions, there are limitations that
hinder its potential [2]. Furthermore, Beer, Clark and Jones [3] state that it is not clear to what extent LMSs
affect students’ participation in activities in universities. These limitations can come in the form of conflicts,
contradictions, miscommunication and misunderstanding in learning systems that can affect students’
participation in e-learning activities. With this focus, using Activity Theory as its research framework, this study
nd
focuses on the contradictions that emerged in the form of tensions, frustrations, misunderstandings and
miscommunication in a fully online university course in New Zealand.The paper includes some of the findings
of this case study and a discussion
ssion on how students’ participation in learning activities facilitated by Moodle
was affected by these contradictions.
II. RESEARCH CONTEXT AND RESEARCH QUESTIONS
The course which was the focus of this case study was one of the papers of the Graduate diploma of
Teaching and was taught fully online for 12 weeks in a university in New Zealand. All class interactions took
place in the university LMS, Moodle, and most of
of the activities carried out in this paper were asynchronous. The
participants who took part in my research were three students (Irene, Hannah and Jake) and their lecturer(Laura)
from one stream out of three. The lecturers of other two streams (Faye and Michelle)
Michelle) also participated in my
research. The lecturers of these three streams (Laura, Faye and Michelle) worked as a team but had their own
ways of communicatingin their own streams. For example, the lecturer of the group under this case study,Laura,
uploaded
aded voice files at the end of each week as feedback to her students’ weekly discussion forums.
The research questions that guided the data analysis of this research were:
•
What were the contradictions that emerged within an activity system of a fully online
onlin
university course facilitated by Moodle?
•
How was students’ active participation in the course affected by these contradictions?
III. ACTIVITY THEORY AND ITS PRINCIPLE OF CONTRADICTIONS
Activity Theory is derived from socio-cultural
socio
and socio-historical theories and through
hrough the lens of Activity
Theory, learners’ constructionof knowledge can be observed and analysed explicitly. Activity Theory can offer
insights into learning processes that are facilitated by Learning Management Systems, in particular, it allows
scholars to examine and document successful and unsuccessful incorporation of technologies and activities that
are facilitated by LMSs in online learning environments.
Engeström’s[4] Activity theory framework offers analytic tools that are appropriate for
f modelling activity
systems. The constituents of an activity system include subject, object, tools, rules, community and division of
labour. Figure 1 shows the basic structure of Activity Theory.
Figure 1: Activity Theory framework (adapted from Engeström[4])
nd
International Conference on E-Education
E ucation & Learning Technologies ( ICEELT 2013 ) ,
Kaptelinin and Nardi [5]; Engeström [6]; Kaptelinin, Nardi and Macaulay [7] and Kaptelinin[8] have
discussed the characteristics or principles of Activity Theory. Some of these basic principles of Activity Theory
include object-orientedness, hierarchical structure of activity, internalization vs. externalization, mediation,
development, multi-voicedness of activity systems, and contradictions as a source of change and development,
which is to the focus of this paper.
Contradictions, as a basic principle of Activity Theory assist in identifying the tensions and conflicts that
emerge in systems of online learning environments.Contradictions are also referred to as structural tensions that
may have accumulated over time. These contradictions may create conflicts, interruptions and clashes: however,
through the resolution of conflicts, they can also be considered as sources of change or development.
Contradictions are defined as “a misfit within elements, between them, between different activities, or
between different developmental phases of a single activity” [9]. Engeström[4] proposes four levels of
contradictions (1) primary, (2) secondary, (3) tertiary and (4) quaternary. The primary contradictions occur
within the elements of activity systems (e.g. within the community). Secondary contradictions arise between the
elements of an activity system (e.g. between the community and subject), tertiary contradictions arise when
activity participants face situations where they have to use an advanced method to achieve an objective (e.g.
when they are introduced a new technology), and quaternary contradictions occur between the central activity
system and outside activity systems. In the context of my research, the contradictions that emerged within and
between(primary and secondary) the elements of the activity system are illustrated in this paper. The following
figure demonstrates a synchronous activity (forum discussion facilitated by Moodle) that is placed within the
Activity Theory framework.
Tools
Moodle, computer,
Object (critically
reflect and critique
the topic questions
and key ideas from
literature)
Subject (student
participants)
Outcome:
Applicable
knowledge
Division of labour (lecturer’s
voice files to individual groups,
Rules (APA referencing
style, word limit, 3
contributions per week)
Plenary voice files to all)
Community (peers,
lecturer)
Figure 2: Asynchronous forum discussion activity system
nd
IV. RESEARCH METHODS AND DATA ANALYSIS
The methods used for data collection of this research included interviews, observation of online activities
and document analysis (course outline) and as a data management tool, NVivo was used. For the purpose of this
paper, two asynchronous learning activities:
(1) PowerPoint presentation prepared for a conference and a reflection uploaded on Moodle, and
(2) Weekly forum discussions that were facilitated by Moodle, were selected for the data analysis.
Engestrom’s[4] Activity theory model was used to organize the findings to identify the sources of these
tensions and conflicts within and between elements of activity systems. In coding the data, relevant meaningful
units from transcribed interviews, observational notes and relevant documents were identified and then were
coded according to the elements of Activity Theory as a method of typology. Among the sub-themes that
emerged under these main themes or the elements of Activity Theory, contradictions became visible in a variety
of forms. These contradictions are illustrated in the following section.
V. FINDINGS AND DISCUSSION
The findings revealed that several contradictions occurred within and between the elements of activity
systems in this context. These contradictions include issues related to course materials as tools, grading,
communication; and opinions related to ‘teacher’s presence’.
A. Presentation of materials
The teaching materials in this course included journal articles, You Tube clips, PowerPoint slides and
Podcasts. Students’ voices indicated the way they felt about the presentation of these physical toolsin this course
and these views described the tensions and frustrations they experienced in this course.
Hannah pointed out that the students preferred to have the reading materials in the form of a book, as it was
convenient for them to read. She mentioned that in a communication space many students raised the issue of
having a book instead of several articles every week. She explained that:
Everybody wanted the book for reading. We all asked for it, and then a lecturer said that we can’t have
it, but afterwards she agreed because the request continued. Now she has agreed to it, but until I see it,
I don’t believe it you know for the next semester. They might say they haven’t. They have to be
organized enough to think of the materials 4 months in advance (interview 2)
This issue was also highlighted by Irene. In her opinion, they could go back to the readings easily if they had
the hard copy of a set of readings. She feels that not having a hard copy of the articles has “actually been a bit of
a pain” (interview 2). However, when Michelle,who is the coordinator of the course,was asked regarding the
issue of giving a printed copy of readings, she had a different view on that. She believed that by giving the
articles every week to students, they could ensure the most recent publications were used.
The process of putting the readings online rather than in a hard copy has worked well because as the
12 weeks have unfolded and new papers have been published that are related to play that are quite new
and exciting, so we can upload and talk about them (Michelle, interview)
nd
Students’ voices suggested that they still preferred to have the book and if the lectures would like to give
them new materials, they can upload extra materials on Moodle. Hannah also pointed out that “Many people
must have printed them all, seriously it’s a lot of money. Students can’t download some stuff (interview 2.)
Having to read on the screen or print articles on a weekly basis and not being able to download some of the
materials was seen as frustrations that created tension in student participants.
In terms of presentation of materials, Irene seemed to have difficulty following some of Laura’s Podcasts
that were uploaded for theirstream. This is because as Irene mentioned, Laura’s Podcasts had background music
in them and Irene, having a hearing difficulty, found it hard to follow them:
One other thing is Laura’s Podcasts, they are often hard to hear because she sets the volume too low
when she is recording. And I’m little bit deaf and it has to be dead quiet for me to concentrate. In one
of them she had music playing in the background it was a great pain (Irene, interview 2)
She also mentioned that she had some issues with downloading Podcasts in this course which affected her
full participation in forum discussions. This was also acknowledged by Irene’s lecturer, Laura. She stated that
she had difficulties downloading the same Podcast and she was aware that students had the same problem:
The students had some difficulties I think getting into the Podcasts. Even I tried it at my home computer
and it wasn’t easy and it took me hours to download that drama thing. That can really preclude full
participation (Laura, interview 2)
Laura’s view suggests that the students’ active participation could have been affected when the students
could not download the podcasts. Jake explained why some of the students could not open some of the Podcasts
uploaded in this course. In his opinion, they were recorded using different formats and some students could not
download them. As a suggestion he said planning and testing should be done before the course starts for a
smooth run:
It seems that some of the Podcasts you could sync with iTune and you could automatically download,
but then the other Podcasts were not through iTunes and it was in different formats, may be it was in
an audio file or just the iTune thing and not everyone has iTune, so they should do just the audio
Podcast and keep it nice and simple. I suppose it’s trying out different things and seeing what’s best.
As we were told we have to be up-to-date with technology. Just realizing that it’s an evolving process,
but still being aware that things have to run smoothly like there’s testing and everything, but should
plan properly (interview 2)
The contradictions that emerged within the element Tools in the activity system were in the form of
frustrations and difficulties,and these seemed to hinder students’ participation in some ways.
B. Issues on grading
In their assignment one, drawing on to the discussions and the literature, students were to create a
PowerPoint presentation for a conference and also write a 1000 word reflection based on a set of reflective
questions given and uploaded on Moodle. However, students were not sure whether they were to include notes
in the PowerPoint presentation and some of the students posted questions on the Moodle communication space
asking for clarifications from their lecturer, Laura. When students got a reply from Laura to the question, the
nd
students realized that they had been given contradictory information.Laura asked them to include notes while
Faye, the lecturer of another streamhad asked them not to include the notes in a previous conversation. In
Irene’s point of view, one person should have given instructions regarding assignments:
There was confusion there. Laura actually gave us wrong information. I think one person should be the
person that deals with the actual sort of what is required for the assignments(Irene, interview 2)
Students also seemed to be frustrated when thisassignment was marked by someone else. Laura mentioned
in an interview that she did not mark the first assignment—the PowerPoint slides they prepared for a conference
and the reflection on that, as she was away at a conference in the United States. She mentioned, “I didn’t in the
end mark the first assignment because I was away, but I did have a look at a few and I was very pleased with it”
(interview 2). However, related to the marking of assignment one, Hannah expressed her concern and stated that
she is worried that someone else is marking their assignments.
At this moment my frustration is I know for sure that they are not correcting my assignments.
Somebody else is correcting. It’s like I got the name of somebody else who has corrected my
assignment (Hannah, interview 2)
When Laura was asked whether she would get help with marking students’ assignments, she assured that she
was going to mark them on her own. However, the students were not informed of any of the arrangements or
plans regarding the marking.
These frustrations and tensions arose between the subject (students) and the community (lecturers) in this
context as a result of not having thorough planning in terms of the lecturers’ roles and responsibilities (division
of labour). Hannah’s worry of someone else marking her assignments could be avoided if the students were
given clear information on why Laura did not mark the first assignment and an assurance that she would
moderate them and would mark the rest of the assessments.
C. Issues on communication
Laura’s absence, and someone else marking students’ assignments, further complicated issues when students
tried to communicate with Laura assuming that she was marking their assignment one. Irene described that after
she submitted her assignment, she realized that she had forgotten to include references, so she sent a message to
Laura in the one-to-one space on Moodle asking whether she could send the references. Since Laura was away,
Michelle looked after Laura’s stream and got access to Laura’s one-to-one space. Irene was not aware of this
arrangement and also the fact that someone else marked her assignment. Irene said:
I sent a note on Moodle in one to one space saying could I resend the assignment and Michelle sent a
reply saying yes just send it, but I don’t think Laura got it because the feedback says that it’s with
incomplete references(Irene, interview 2)
Irene believed that although Michelle replied giving permission to send the references, her message or
references did not seem to reach Laura. This was seen as a gap in their communication link that created
frustration.Prior to this incident Laura also experienced an issue that she was locked out of the one-to-one
communication space. She explained how she got to know that she was not able to see students’ messages
because of the way (assigned role) she was brought into the course.
nd
The other problem I had was that in the one on one space I assumed that I hadn’t heard from any of my
students in the whole course. And then, quite recently, a student emailed me and said “you haven’t
answered my question on one on one space”. When I tried to go in I discovered I had been locked out
of it, so I had no way of knowing….I think it was because when I was initially brought into the course I
was brought in on the wrong criteria (Laura, interview 2).
In relation to communication issues, Hannah also had an experience where her message was not answered
for some time. Hannah wrote to her lecturer and asked for an example of an art work so that she could get an
idea of exactly what she needed to do. This may be due to Laura’s issue of being locked out of the one-to-one
space. Hannah mentioned that when she didn’t get a reply from them, out of frustration she started making her
own art work which was quite different:
I asked them to send me an example before the two week holiday because that was really a good break
you know. Then I didn’t get a reply for a very long time. That was a drawback because I somehow
started doing something that was completely different, but anyway thank god that there was no right or
wrong, so I submitted it. But you know when we ask something…it’s not like we are desperate at least if
we get that…I mean this is the least we get back from our teachers (interview 2)
D. Different student-lecturer and lecturer-lecturer opinions on ‘teacher’s presence’
The lecturers in this course did not participate in weekly forum discussions, but they uploaded feedback in
the form of voice files and plenary podcasts in most of the weeks. This created controversial opinions between
students and lecturers as well as among lecturers. Faye attested that, “our philosophy in the program is not
going to the discussions during the week because it does interrupt the flow of what people are saying”
(interview 1). Michelle shared similar views regarding “teachers’ presence” and she explained that if the
lecturers are there, the students may not freely express their opinions. Michelle commented:
I’d much prefer that we go in at the end or beginning in the week to do the voice file…… I’m not going
to intervene in the conversation because I know that would shut it down and students will respond
thinking “she’s the lecturer and we better say that”. I think it allows them more opportunity to have
honest discussions with each other(interview)
On the other hand, students felt that they were left on their own and they were not sure whether they were on
the right track without the “teacher’s presence”. The students seemed to need some kind of guidance and
acknowledgement of their contributions. For example, Hannah mentioned:
I did another course and the lecturer used to be part of our online discussion and lead us through it.
That method is better because you feel the presence. Your teacher is there…. But here they say happy
discussion and they post it…I don’t know, for example this is the end of 3rd week and I haven’t got any
feedback what I have been doing, so I really don’t know whether I’m doing the right thing (interview
1).
Regarding ‘teacher’s presence’ in this context, Laura had a different opinion to Faye and Michelle, as she
was more keen to have a dialogue with her students than giving them a talk at the end of the week. She
accentuated:
nd
Well as I said the discussion voice files ….my voice files to the students were my way of having a
genuine dialogue with them when I couldn’t go into their online space, so I think it would be best to
establish a way of having that dialogue as a reciprocal exchange - not just mere responding as an end
point, like a plenary (Laura, interview 2)
Laura believed that two way communication is important for her, as that’s her way of having a dialogue with
her students. Laura said, “Dialogue to me is not “you speak and then I respond”, it’s an on-going reciprocal
thing. I think that I would like to have more of that in this course”(interview 2).She also mentioned thatthere
were times in the discussions she would have quite liked to go in and steer it in a different direction. However,
when Laura was asked whether she meant that the lecturers should participate in forum discussions, she was not
sure whether that was what she wanted.Alternatively she suggested that if she had a choice she preferred not to
grade all the discussions, but just a few of them and let the students have more free thinking:
I think in a way, if it was my choice, I would prefer not to grade but instead create a kind of capacity to
‘free fall’ as I call it...free fall thinking. It’s somewhat constrained, with the grading. May be I wouldn’t
grade all the discussions. I might grade perhaps the last three or various parts rather than all, but it’s
not up to me (interview 2)
The differing opinions on “teachers’ presence” in this context denoted clashes between students’ and
lecturers’ views. Students’ views demonstrated their need to have the lecturer as part of their community while
engaging in forum discussions. With lecturers’ absence in forum discussions, the findings alsoshowedan
absence of reciprocal communication between the students and the lecturers that interrupted students’
participation.
VI. CONCLUSION
As Kuuti[9]suggests, contradictions can be “problems, ruptures, breakdowns, clashes” in activities. The
contradictions that emerged within and between the elements of activity systems in this case study were seen in
diverse forms such as frustrations, tensions, difficulties and contradictory opinions. Presentation of course
materials that included journal articles and Podcasts as physical tools created tensions among students when
they were unable to download some of them and when they had to read on the screen or print them each week.
The lack of planning in terms of individual responsibilities (division of labour) also caused confusion in this
context when the students were given contradictory information. As suggested by the students, lecturers’ roles
as well as information provided regarding assignments and marking should have been explicit. The
contradictions related to the issue of miscommunication was also due to lack of planning and communication.
As a result of bringing Laura under the wrong category (assigned role) in Moodle, both the students and their
lecturer were frustrated. The opposing views on lecturers’ participation in forum discussions indicated the
clashes that were in this activity system that affected students’ participation in this course. Lecturers’ feedback
and guidance as well as reciprocal communication in forum discussions were rather limited since the lecturers
decided not to be part of the community in this context. Therefore, successful implementation of an online
course that is facilitated by an LMS needs careful planning that suits the needs of students. This includes
structural strategies in the design of the course. The findings from this research may help overcome tensions and
contradictions when designing courses and activities in online learning environments.
nd
References :
[1] H. Coates, J. Richard, and G. Baldwin, “A critical examination of the effects of learning management
systems on university teaching and learning,” Tertiary Education Management, vol. 11, no. 1, pp. 19-36,
2005.
[2] W. R. Watson, andS. L. Watson, “An argument for clarity: What are learning management systems, what
are they not, and what should they become?,” TechTrends, vol. 51, no. 2, pp. 28-34, 2007.
[3] C. Beer, K. Clark, and D. Jones, “Indicators of engagement,” In Curriculum, Technology and
Transformation for an unknown future, C. H. Steel, M.J. Keppel, P. Gerbic and S. Housego, Eds., Sydney:
Proceedings ascilite Sydney, 2010, 75-86.
[4] Y.Engeström, “Learning by expanding: An activity-theoretical approach to developmental research,”
Helsinki: Orienta-Konsultit, 1987.
[5] V. Kaptelinin, and B. A. Nardi, “Acting with Technology,” Cambridge, MA: The MIT Press, 2006.
[6] Y.Engeström, “Expansive learning at work: Toward an activity-theoretical conceptualization,” Journal of
Education and Work, vol. 14,pp. 133-156, 2001.
[7] V. Kaptelinin, B. A. Nardi, and C. Macaulay, “The activity checklist: A tool for representing the "space" of
context,” Interactions, vol. 6, no. 4, pp. 27-39, 1999.
[8] V.Kaptelinin, “Activity theory: Implications for human-computer interaction,” Context and Consciousness:
Activity Theory and human-computer interaction, B. A. Nardi,Ed. Cambridge, MA: The MIT Press, 1996,
pp. 103-116.
[9] K.Kuutti, “Activity theory as a potential framework for human computer interaction research,” Context and
consciousness: Activity theory and human computer interaction, B. A. Nardi, Ed. Cambridge: MIT Press,
1996, pp. 17-44.
nd
Rapid Easy Authoring Platform for Serious Games A Game-based Learning System
for Work Safety Training
Peter Leong
Singapore Polytechnic, Singapore
[email protected]
Teo Kim Sheng
Singapore Polytechnic, Singapore
[email protected]
Abstract:
In recent times, games-based learning has been attracting the attention of educators around the world.
One of the challenges in deploying games-based learning to cover a larger proportion of the curriculum is the
high costs associated with development of the games. Games development often requires specialized skills sets.
Inputs from academic specialists for the learning content and pedagogy is also critical to the successful
development of games-based learning. Our REAPSG platform empowers academics and trainers to create the
games-based learning content on their own without the need for specialized development skills.
Keywords-component; game authoring tool; reusable game components; game templates; easy authoring;
game knowledge mining; game evaluation
I. Introduction
The rapid proliferation of computer games on the internet and on smart mobile devices have enabled
computer games to become a new media for mass communications. Many students in our institutions of learning
spend a significant number of waking hours playing computer games. Many educators have experimented with
games-based learning in class [1][2][3] to increase student engagement and motivation in learning
[4][5].However, the high-costs (commonly reported in the millions of dollars) of creating customized game
content for a particular lesson and curriculum have deterred wider adoption of game-based learning [6][7].
The high-costs of creating customized games stem from the fact that you require specialized 3D artists,
modelers and animators in the game creations process. Specialist 3D graphics, physics, artificial intelligence and
network programmers are also involved in the game creation process. Lastly, games created for a particular
curriculum or geography often have a limited market size, and hence there is an insufficient number of users to
spread the costs of game development.
For specific content domains such as work safety and health, the game developers need inputs for the
domain experts in order to maintain the authenticity of the game setting and scenarios. Such domain experts
may havelittle understanding about games technology. Hence, the frequent communications between the domain
experts and game developers is necessary and this also drives up the costs of development; that is, the costs of
engaging the domain experts as consultants in the project has to be included in the total project costs.
Our proposed approach to address this problem of high costs of game production is to build an easy-touse set of tools such that academics and domain experts can create their own game-based learning content [12]
nd
using our tools and deploy them to their students through the web browsers. The cost of creating the generic
platform which we have named REAPSG (Rapid Easy Authoring Platform for Serious Games), can be factored
over large number of potential users who will all use the platform to create content. The REAPSG platform will
also provide some pre-configured game templates to assist the academics and lecturers in preparing the gamebased learning content for their lessons.
II. Related Works
A number of projects have been initiated to use games-based learning in the classroom. The project [8]
in Malaysia reported successful deployment of games-based learning among secondary one students. In
Singapore ,games-based learning has been used in schools [9][10][11]. In Victoria University, Australia, they
have been using games-based learning for vocational training in the construction industry [12]. One source of
the game based learning content for this projects have been commercial-off-the-shelf games repurposed for an
educational outcome. Other projects have a customized game created specifically for their educational purpose,
and end-users and trainers are not able to add to the content or modify it significantly. Some games-based
learning projects make use of virtual world environments such as Second Life [13]. However, construction of
new content logic within Second Life require specialized skills which are not easily picked-up by general
academics and lecturers. Our approach in REAPSG is to make it more limited in functionality, but to make it
very easy for any general academic or lecturer to be able to use the platform with minimal training.
III. Overview of the Rapid Easy Authoring Platform for Serious Games
The REAPSG architecture is based on the concept of empowering the teacher to create their own
game-based content [14][15]. Teaching and engaging educators about serious games is challenging, but it is a
model that can scale-up to many more courses and students with the engagement of educators in the process.
Our REAPSG architecture must allow educator to create their own customized games [14][15][16], should
mitigate the difficulties educators face in using game engine tools by themselves [17], provide some jump start
content (prefabricated content templates to speed up the process), provide libraries of easily re-usable assets, and
it should not require computers with high 3D graphics processing capability.
A. Architecture
The system is divided into two components: 1) the game authoring tool or game editor, and 2) the game
player component. The game states during play are recorded in a database on the game server. In order for
student or learner to play the game, they must use a web browser to download the browser plug-in or web
player. This is done automatically when the user opens a particular URL that points to the lesson content.
The game authoring tool or game editor (see figure 1) is the main tool for academics and subject matter
experts to create the game content. It runs as the desktop client on the lecturer’s workstation or laptop. After the
lecturer has finished editing the game-based lesson, the lecturer would save the game state and it is recorded in a
database on the game server. The REAPSG provides a number of basic game templates such that the lecturer
does not have to create the entire game world from the ground up if he or she wants to just modify the content to
suit his own purpose. However, if the lecturer wants to create everything in the scene from a blank slate, he
would also be able to do so.
nd
The game player component for REAPSG runs in a web browser (see figure 2). In order for student or
learner to play the game, they must use a web browser to download the game at a particular URL that points to
the lesson content. The URL link is specially encoded with information that points only to a particular lesson.
Authentication of the user can occur via normal web-based authentication schemes.
If a web-based lesson management system (LMS) such as Blackboard is available, the URL based
game loading mechanism can easily be integrated with the LMS content. It is the lecturer/trainer who must use
the game editor tool to create the game-based learning content, and when he saves the edited work on the server,
it will also generate an encoded URL which is linked to the game server content. The lecturer must copy the
URL string (as generated) into the LMS (e.g. Blackboard) page. The URL mechanism is portable across all
web-based LMS as it is an inherent part of the WWW cyber-infrastructure.
Figure 1. Virtual construction yard created using the REAPSG game authoring tool/editor
One of the concerns for 3D content is the size of the file to be down-loaded via the communications
link. We have optimized the system to reduce the amount of data to be retrieved to a minimum. We actually
download the 3D model, geometry and textures instead of video format of the game scene, and the rendering is
done using the graphics processing capability of the user’s own laptop or workstation; this reduces the size of
the data significantly. The corollary concern is the graphics processing power of the laptop or workstation and
whether it will be able to render the 3D scenes with sufficient quality at an acceptable speed. Most modern
laptops and workstations have basic 3D rendering capability, so if the game 3D models are kept to low-polygon
count, the rendering time would be reasonable. We may not be able to support some of the 3D effects such as
anti-aliasing, fog and atmosphere on machines without better 3D graphics cards. The rendering engine being
used for REAPSG must be able to right-size the quality of the 3D content according to capabilities of the
hardware; for example, HD quality graphics on high-end hardware but lower resolutions on lower-end
hardware.
We have also partitioned the content into separate asset bundles such that you only down-load the 3D
graphics models you are using in the lesson instead of all available models in the entire REAPSG library. Aside
from minimizing download times for the web player, such an approach will also make the system easy to
expand in the future, as new assets can simply be added to the server without requiring large amounts of code
changes.
nd
Figure 2. Player view of the construction yard within a web browser
B. Implementation
To enable rendering of the graphics on multiple platforms ranging from desktop computers, to web
browsers and mobile computing devices, we decided to use a game middleware. We selected the Unity3D game
middleware because of its wide support of different devices playback platforms. Unity3D is able to support the
Flash platform, iOS, Android, Windows desktop, Mac OS desktop, Wii U, PS3 and Xbox360. The Unity3D
provides the graphics rendering capability, visual effects such as fog and particle effects, and also the 3D
physicsfor game objects such as friction, elasticity, acceleration, velocity and angular momentum. The game
physics engine does not support a number of physics effects such as fluid flow, thermodynamics, convection,
conduction and electricity. Such physics effects, if needed, must be implemented in code by the programmer in
the REAPSG platform.
The REAPSG supports two basic types of game formats. The first is a “Spot-the-X” format where the
player has to use visual cues to distinguish items which have something amiss from the normal items in the
scene. The challenge in this game format is largely based on time and ability to pick up visual cues using
content knowledge about the setting of the game scene. For our REAPSG, our initial target application is
construction work safety.For the “Spot the-X” format games, the safety trainees must use their knowledge about
safety to indicate which conditions are hazardous in the scene. For example, a scaffolding without proper means
of securing it would be hazardous. The safety trainees must be able to complete a level and spot a variable
number of hazardous conditions within a time frame. The game will be played from a first-person perspective,
to provide greater immersion and realism for the player as he/she navigates in the virtual worksite looking for
potential hazards.
The second format of games that REAPSG supports is the “Branching-Story” format. This game
format represents a scenario as a tree of possible outcomes given different actions. Based on the trainee
sequence of input decision, he will traverse along a different branch or path through the tree. Each different path
or branch will lead to a different outcome; some successful, some partially successful, while others end in
failure. The “Branching-Story” allows the trainer to capture repercussions of a series of less than optimal
decisions. It is more flexible but less visually oriented compared with the “Spot-the-X” format. The REAPSG
uses still images as the backdrop for the scenarios in order to create some amount of interest and to create a
nd
setting for the story elements in the “Branching-Story” game. For example, instead of purely using text to
describe a construction machine, we will be showing the machine in an image form with text descriptions of the
scene.
Figure 3. Branching stories view in the REAPSG player
The lecturer or trainer uses the REAPSG game authoring tools to create both “Spot-the-X” and
“Branching-Story” games. The completed game scenario is stored on the server. When a trainee launches a
game on his browser, the information about the game scenario is downloaded to the web browser and displayed.
Figure 4. Question & answer screen in the REAPSG player screen
Our current implementation is based on broadband or LAN speeds between the game server and
browser (it will not work over modems or GSM/GPRS connections). The player component on the web-browser
can execute both “Spot the-X” and “Branching-Story” games.
IV. User Evaluation of the Rapid Easy Authoring Platform for Serious Games
Getting feedback from the lecturers and trainers who have to use the REAPSG to create games-based
learning content and getting them to actually use them in their lessons is a critical success [18] factor in the
whole project. Our subject matter experts in Work Safety and Health (WSH) are not very familiar with gamesbased learning. However, we have created a platform for them to share their views with the development team
during the course of the project.
nd
To improve the brain storming process, all the game designers and developers attended a design
thinking workshop prior to the actual commencement of the project. Using design thinking, we hope to achieve
deep user empathy with the needs of the WSH trainers. We also used user observations as the WSH trainers
used the initial prototypes of our system to gauge their reaction and to understand their viewpoint as they use
such as gamesbased learning system for the first time.
To aid our decision and design process, we have also created a score card [19][20] or rubric on the
features of the system when it is matched against the WSH trainer voiced and unvoiced requirements (see table
1). Each trainer is allowed to have only a fixed number of vote which he can put into any cell, or put all into the
same cell.The row score indicates the usefulness of including that feature in REAPSG. The column total
indicates the overall number of matches in each of the mandatory, good and optional categories. The total row
score is an overall score for the system, a higher value indicates we have are closer to deriving maximum value
for the REAPSG system for the trainers who will be using it.
TABLE I. SCORE CARD FOR REAPSG
V. Conclusions
The REAPSG is a system that allows subject matter experts and lecturers to create their own
customized games-based learning content easily, quickly and affordably. In the future, we are planning to
expand the number of game templates created and also pilot the system with a number of students in the
Diploma in Work Safety and Health (WSH). We believe REAPSG will be a step forward towards more
authentic learning in WSH and will lead to greater knowledge and retention by the trainees. We also believe the
interactive content will increase the level of engagement and motivation of the students.
ACKNOWLEDGMENT
This project is made possible with funding support from the Institute of Adult Learning CET
Innovation Fund. Together with subject matter expertise from The Professional & Adult Continuing Education
(PACE) Academy, Singapore Polytechnic.
nd
REFERENCES
[1] Prensky, Marc. "Digital game-based learning." Computers in Entertainment (CIE) 1, no. 1 (2003): 21-21.
[2] Van Eck, Richard. "Digital game-based learning: It's not just the digital natives who are restless."
EDUCAUSE review 41, no. 2 (2006): 16.
[3] Papastergiou, Marina. "Digital Game-Based Learning in high school Computer Science education: Impact
on educational effectiveness and student motivation." Computers & Education 52, no. 1 (2009): 1-12.
[4] Ebner, Martin, and Andreas Holzinger. "Successful implementation of user-centered game based learning in
higher education: An example from civil engineering." Computers & Education 49, no. 3 (2007): 873-890.
[5] Foreman, Joel. "Game-Based Learning: How to Delight and Instruct in The 21st Century." Educause Review
39, no. 5 (2004).
[6] Torrente, Javier, Á. Del Blanco, Eugenio J. Marchiori, Pablo Moreno-Ger, and BaltasarFernández-Manjón.
": Introducing educational games in the learning process." In Education Engineering (EDUCON), 2010
IEEE, pp. 1121-1126. IEEE, 2010.
[7] Oblinger, Diana. "Games and learning." Educause Quarterly Magazine 29, no. 3 (2006): 5-7.[Online].
Available: http://www.educause.edu/ero/article/games-and-learning.
[8] Eow, Yee Leng, and RoselanBaki. "Form one students’ engagement with computer games and its effect on
their academic achievement in a Malaysian secondary school." Computers & Education 53, no. 4 (2009):
1082-1091..
[9] Tim Marsh, Chuang Xuejin, Li ZhiqiangNickole, Scot Osterweil, Eric Klopfer, and Jason Haas. 2011. Fun
and learning: the power of narrative. In Proceedings of the 6th International Conference on Foundations of
Digital Games (FDG '11). ACM, New York, NY, USA, 23-29.
[10] Mark J. Chavez, Andrew Gregory, Lau Yeung, and Wang Xueli. 2009. Voyage to the age of the dinosaurs
project, an experiential learning situation with undergraduates, graduates and visiting professionals. In
ACM SIGGRAPH ASIA 2009 Educators Program (SIGGRAPH ASIA '09). ACM, New York, NY, USA, ,
Article 6 , 4 pages.
[11] Henry Been-Lirn Duh and Vivian Hsueh-Hua Chen. 2010. Fantasies in narration: narrating the
requirements of children in mobile gaming design. In Proceedings of the 7th International Conference on
Advances in Computer Entertainment Technology (ACE '10). ACM, New York, NY, USA, 3-6.
[12] O'Rourke, Mark, Juanita Custance, and TasPapasimeon. "Developing a Vocational Training Computer
Game
Workplace
Simulator:
The
Vocational
Game
Project."
(2008).
[Online].
Available:
http://vuir.vu.edu.au/15867/1/mark_orourke_vocgame.pdf
[13] Tassawar Iqbal, Klaus Hammermüller, and A. Min Tjoa. 2010. Second life for illiterates: a 3D virtual
world platform for adult basic education. In Proceedings of the 12th International Conference on
Information Integration and Web-based Applications & Services (iiWAS '10). ACM, New York, NY,
USA, 373-380.
nd
[14] Becker, Katrin. "Digital game‐based learning once removed: Teaching teachers." British Journal of
Educational Technology 38, no. 3 (2007): 478-488.
[15] Becker, Katrin. "Teaching Teachers about Serious Games." In World Conference on Educational
Multimedia, Hypermedia and Telecommunications, vol. 2007, no. 1, pp. 2389-2396. 2007.
[16] Squire, Kurt, Levi Giovanetto, Ben Devane, and Shree Durga. "From users to designers: Building a selforganizing game-based learning environment." TechTrends 49, no. 5 (2005): 34-42.
[17] Dickey, Michele D. "Ninja Looting for instructional design: the design challenges of creating a game-based
learning environment." In ACM SIGGRAPH 2006 Educators program, p. 17. ACM, 2006.
[18] Blunt, Richard. "Does game-based learning work? Results from three recent studies." In The
Interservice/Industry Training, Simulation & Education Conference (I/ITSEC), vol. 2007, no. -1. National
Training Systems Association, 2007.
[19] De Freitas, Sara, and Martin Oliver. "How can exploratory learning with games and simulations within the
curriculum be most effectively evaluated?." Computers & Education 46, no. 3 (2006): 249-264
[20] Dondi, Claudio, and MichelaMoretti. "A methodological proposal for learning games selection and quality
assessment." British Journal of Educational Technology 38, no. 3 (2007): 502-512.
nd
The Impact of Web Project-Based Learning on Elementary School Students’ Development of
Van Hiele’s Geometric Thought in Taiwan
Tzu-Hua, Huang
Department of Education,
Taipei Munipical University of Education, Taipei, Taiwan
[email protected]
Yuan-Chen Liu
Graduate School of Curriculum and Instruction, and Communications and Technology,
National Taipei University of Education, Taipei, Taiwan
[email protected]
Ru-Fen, Kuo
Graduate School of Curriculum and Instruction, and Communications and Technology,
National Taipei University of Education, Taipei, Taiwan
[email protected]
Abstract
The Dutch mathematics educators Van Hiele couple in 1957 proposed a model of theory described how
children develop geometric theorems, which consisted of five levels of understanding: visualization, analysis,
informal deduction, formal deduction and rigor. The development of the levels is more dependent on the
instruction received than on children’ age of maturity. Learners advanced from basic level (visualization) to the
highest level (rigor) through instruction and learning experience. Therefore, whether students can achieve high
level of understandings, curriculum design and teaching methodology are the significant factors. Project-based
learning enhances students’ learning motivation, high-end level of problem-solving and thinking skills, which is
one of the systematic teaching methods. Students involve in activities that integrated knowledge and learning
skills through the process of exploring complex authentic- questions and carefully designed tasks. With the
development of technology, project-based learning gradually tends to be web-based. The study adopted Web
project-based learning to fifth to sixth grade students in elementary school to explore their learning
effectiveness and development level in Van Hiele’s geometric theory. The result revealed that students’
performance through Web project-based learning has improved on level one and level two of Van Hiele model.
Keywords:Web project-based learning, Development of geometric thought, Geometric concept
I. Introduction
A married couple of Dutch mathematics educators, Pierre van Hiele and his wife, published companion
dissertations in 1957, describing a concept that there were five levels or stages of Geometric Thought that any
one student could be at, and the level of development was related to teaching factors, it was not influenced by
the children's age and maturity; having an appropriate teaching process provided by the teachers, students would
sequentially experience these five levels (level 0-4) [13]. Additionally, Crowley also pointed out that van Hiele
nd
believed that the growth occurred at various levels was primarily subjected to the "teaching", rather than the
maturity along with an individual's age, therefore, the organization and methods of teaching, the selection and
use of text books was very important [11]. Therefore, teachers needed to constantly seek the innovative teaching
methods to motivate their students in order to improve learning outcomes for all students [17] [12]. In the past, it
used to focus on specialized mathematical knowledge and structure, basic computing skills, and learning to
solve routine mathematical problems etc. and it had now turned to emphasize the concept that students must
actively construct mathematical knowledge [5].
The Buck Institute for Education located in the Northern California, United States, continuously
promotes the Project-Based Learning(PBL) as their organization's objectives, and they define PBL as "A
systematic teaching method, through the complex and real problems combining with a carefully designed
mission exploring process, which can make students to be integrated into the knowledge and skill learning
activities”. Over the years, many research literatures showed that PBL indeed had many positive educational
outcomes [9] [2], in particular, it could contribute to improve the students’ learning motivation, high-level
problem-solving skills and the ability to think, enhancing the professional knowledge, as well as creating a
positive classroom atmosphere, improving cooperation skills, so as to drive the students to actively learn for
themselves and learning interests, and hence to enhance the students' knowledge, skills and affective and
problem-solving ability. Thus, the Project-Based Learning(PBL)combining with Internet Technology is a
successful teaching model, it will not only increase students' interest in learning, but also it can activate the
students' critical thinking skills, therefore, the purpose of this study is planning to utilize the Web-based and
PLB S system combining with the geometry course in existing mathematical textbooks to design the teaching
methods, and to explore students' growth and change in the levels of geometric thought.
II. Literature Review
Van Hiele’s Geometric Thought
In 1957, a married couple of Dutch mathematics educators, Pierre van Hiele and his wife, proposed a
concept that there were five levels or stages of geometric thought that any one student could be at, and the level
of development was related to teaching factors, it was not influenced by the children's age and maturity; having
an appropriate teaching process provided by the teachers, students would sequentially experience these five
levels, which comprised: Visualization, Analysis, Informal Deduction, Formal Deduction, and Rigor [7].
According to these theories, the construction of geometric knowledge would sequentially go through five levels
and each level had its own very unique characteristics [7]. According to Crowley, the van Hiele’s levels of
geometric thought had contained five characteristics, which comprised sequential, advancement, intrinsic and
extrinsic, linguistics as well as mismatch [11].
Van Hiele insisted that students' learning progress between all levels should rely on the organization
and methods of teaching as well as the selection and use of textbooks, instead of relying on the maturity levels
of the students [11]. Therefore, when moving forward from one level to the next level, teaching activities played
a very important role, and the correct teaching activities would get better grades more easily[1]. The van Hieles
defined five phases in the learning process for guiding students from one level to another on a given topic,
comprising the following sequence: Information, Guided Orientation, Explication, Free orientation, and
Integration [13].
nd
Project-Based Learning
PBL was firstly advocated by Kilpatrick, the primary purpose was to stimulate the students' selflearning interest, it believed that such kind of learning method could not only improve the students’ learning
efficiency, but also it could activate the abilities associated with all other aspects[17].Actually, PBL was a kind
of project-organized learning method, it was a type of learning method that had been developed using a
constructivist philosophy, which aimed to prevent the development of an ossified model of knowledge, going
through a series of exploration actions in connection with cooperative learning situations, so that students not
only could learn the problem-solving knowledge and skills, but also could understand how knowledge could be
applied in daily life [8].
In the aspect of areas, the project-based learning was more suitable for using in the field of
mathematics, physics or engineering [14], which allowed students to learn through reflection on doing, that is,
"Learning-by-doing", and be able to learn from their own experiences, a systematic learning method was being
integrated into the learning process in which the students were responsible for their own learning outcomes, so
that learners could have a better understanding of "how to learn" [10].
When combining the PBL with internet technology, it would be able to make PBL to be even more
effective [3]. The Web- and Project-Based Learning (also known as Web PBL, WPBL) is to learn via the
auxiliary functions of digital learning websites on the internet in order to achieve the goal of PBL.
Krajcik, Czerniak& Berger proposed a Project-based Learning Model (PBL) that asking teachers must
make every teaching element contained in a project to be interconnected, so as to establish the goals of
assessment for students, wherein, the stages for designing a PBL teaching plan were as follows: 1. Identifying
concepts, specifying learning performances, and matching to curriculum objectives;2.Develop the driving
question;3.Develop benchmark lessons;4.Develop investigations;5.Develop assessments; and6.Develop a
calendar[6].
These six stages are an iterative process, many procedures will have been carried out repeatedly, until it reaches
the end of special events or problem is resolved.
This study will construct a teaching and learning website based on the Project-based Development
Model proposed by Krajcik, et al., wherein, teachers will design the interface process in accordance with six
elements of the project-based development model, which allows teachers to enter the information required by
relevant courses; as to the students, they will start to study the course contents provided by the teachers,
according to seven stage of the project-based learning introduced by Krajcik, et al., in which, students will be
conducted from the stage of question guiding to move forward progressively from one point to another on the
way to the formal project-based learning[6]. The students will conduct the stages of PBL are as follows:
1.Messing about; 2.Finding information; 3.Asking and refining questions; 4.Planning and designing;
5.Conducting the experimental work; 6.Making sense of data; 7.Sharing your ideas.
The experimental group planned the overall curriculum activities, where teachers and students are still
linked to each other; while doing a project, teachers will assist the students to clarify the basic concepts related
to course objectives by guiding the questions, and students will start from the exploration activities, collecting
information, integrating, analyzing, and the final stage of learning process is to integrate the whole works.
nd
III. Methods
I. Participants
For concerning about the convenience sampling, the present study chose students from two 5th–grade
classes of one elementary school in Taipei, Taiwan to participate in the research, wherein, one class was the
experimental group, the other was the control group, and the experimental group consisted of 31 students from
one class, the control group consisted of 29 students. This school adopted the normal class grouping, so there
was no significant difference in learning ability level of the students in each class; as for the school curriculum,
the computer education was conducted at least once a weeks in the third grade, all students in the experimental
group had learned the basic skills of using computers, such as, typing, learning to run many Windows
applications, word processing, computer graphics, internet access and information collection etc.
II. Research Design
1) Independent variables
a. Experimental group: WPBL + teachers used an interactive whiteboard (IWB) combining with websites to
teach students + students explored questions while learning through PBL with computers; and
b. Control group: using van Hiele’s five-stage learning method + teachers used an interactive whiteboard (IWB)
combining with websites to teach students + students learned the specific subject matter with computers.
2) Dependent variables
Differences between geometric thinking levels.
3) Control variables
a. Starting point for learning behavior
b. Teaching hours
c. Teaching schedule
d. Teaching materials
e. Instructors
f. Teachers’ teaching tools
g. Students’ learning tools
III. Research tools, the reliability and validity
1) PBL-Online website
The present study designed the website with reference to the development of project-based teaching
model disclosed by Krajcik, Czerniak and Berger, wherein, teachers designed the interface process in
accordance with six elements of the project-based development model, which allowed teachers to enter the
information required by relevant courses; as to the students, they could start to study the course contents
provided by the teachers, according to seven stage of the project-based learning introduced by Krajcik, et al., in
which, students would be conducted from the stage of question guiding to move forward progressively from one
nd
point to another on the way to the formal project-based learning[6]. The overall curricular activities were
coordinated with the process of student learning topics, while doing a project, teachers could assist the students
to clarify the basic concepts related to course objectives by guiding the questions, and students would conduct
the exploration activities, collecting information, integrating, analyzing, and the finally to integrate the complete
works, so students could share and discuss the questions in their classroom, teachers and peers could instantly
provide feedback at the same time.
2) Van Hiele geometric concept test
The van Hiele’s geometry concept test was adopted by this study with reference to “Wu-Shey’s van
Hiele Geometry Test for Elementary School Students” proposed by Wu, and it had been modified in a way to be
the pre-test and post-test tools [1]. The tests involved the first three levels of van Hiele theory of geometric
thought, wherein, each level consisted of three basic geometry concepts which comprised the triangle, square
and circle. This quiz consisted of a total of 70 multiple-choice questions (which each have 4answers to choose
from),in which, there were 25 questions for the first van Hiele level, 20 questions for the second van Hiele level,
and 25 questions for the third van Hiele level. As to the expert validity, two course instructors who taught the
experimental group and the control group respectively were invited to participate in this study, and the results
were reviewed and amended by the advising professor. In the aspect of reliability, the reliability was assessed
using the Cronbach's alpha value (pre-test and post-test was .829 and .807), therefore, these results indicated
good reliability.
3) Reflective journals on the teaching activities
After conducting each teaching activity, the researchers would write a reflective journal to record the
teaching activities, which can act as a teaching reference for the second class; according to the experimental
teaching method adopted by this study, the experimental group and control groups were arranged to learn one
specific subject matter on the same day, hence, when applying project-based learning in the classroom, it needed
to comprehensively reflect the teaching- learning situation so the experimental group and the control group, and
to verify whether students should have different reactions subject to different teaching methods. The encoding
method consisted of the first portion of character, whereas the alphabet "R" representing the data category was
the “reflective journal on teaching activities”, and the second portion of number, whereas there were four
numbers representing the month and day, for example: R0330, indicating that it was a reflective journal to
record teaching activities conducted on March 30th.
IV. Data analyses
The present research designed the questions of van Hiele’s geometry concept test by referring to “WuShey’s van Hiele Geometry Test for Elementary School Students” proposed by Wu, D. B., and the test questions
involving the first three levels of van Hiele theory of geometric thought, wherein, each level consisted of 25
questions for the first level, 20 questions for the second level, and 25 questions for the third level[1].The scoring
method was that a score of 1 for a correct answer and 0 for an incorrect answer. All participants received a score
for each van Hiele level according to Usiskin’s grading system; the criterion for success at any given level was
three or four out of five correct responses[15]. For the student who complied with the Level-N, but had not
reached the criterion of Level-N+1, would be assigned to the Level-N. If a student who complied with the
nd
Level-N, but had not passed the criterion of previous Level-N-1, that is, a jumping level, and could be
considered as unqualified. Any student, who did not comply with the criterion of any given level, could be
classified at level 0.The criterion for success at any given level adopted by this study was four out of five
correct responses.
Meanwhile, an independent-samples t-test was conducted to compare the pre-test scores of students for
two groups in order to evaluate whether there was difference between two groups of students. After completing
the teaching experiment, another independent-samples t-test would be conducted to compare the post-test scores
of students for two groups so as to evaluate whether there was difference between two groups of student in
levels of geometric thought subject to different teaching methods.
V. Course Design
In this study, the course design was performed with reference to general mathematics textbooks sold on
the market. The teaching method for the experimental group adopted was PBL proposed by Krajcik, Czerniak&
Berger, and teachers designed the courses in accordance with six-stage teaching process revealed by Krajcik,
Czerniak& Berger [6]. The course design for the control group was performed based on the five stages of
learning model introduced by van Hiele [13].
a) Teaching contents
The teaching contents used in this study contained ten themes;the course ran for 10 weeks with one
theme per week, wherein, each theme unit was asfollows:Theme1: Preview plus warm-up; Theme2: Triangle;
Theme3: Parallelogram; Theme 4: Trapezoid; Theme5: Kite; Theme6: Rhombus; Theme7: Circle and sector;
Theme8: Thematic integration; Theme9: Presentation; Theme10: Summary and review.
b) Project theme
The project theme used in this study was “Park planning and design”. Thematic contents were as
follows Taipei City Government has a piece of land, in which, there is a rectangular block planning to be used
for a park, although the length, width has not yet been determined, but it is within a range of 1000meterss. The
park is expected to have a small stream, a small bunker, a play yard and a skating rink for children to play,
wherein, it must also contain at least two pedestrian walkways and a gardening area. But it should not end up
designing a dull park area, only one area can have a rectangular shape, and the remaining areas in the park must
be either polygon or circle except a rectangle. The design team should have the talent to come up with creative
ideas. Please draw a design drawing and clearly marked the side length, angle, and area of each region to
facilitate the project construction. At the end of their studies, students must complete the specific project and
make presentations for their works.
IV. Results and Discussion
1) Average number of people who passed the pre-test and post-test of van Hiele’s Levels of Geometric Thought
and its passing rate
According to the results of van Hiele’s geometry test, and when exploring the assessment criterion
which considering the criterion for success at any given level could be four out of five correct responses ,in
addition, this study compared the differences in number of people who passed the test for each level of van
nd
Hiele’s geometric thought and its passing rate. Thus, the number of people who passed the test for each level of
van Hiele’s geometric thought and its passing rate for the experimental group and the control group were as
shown in Table 1.
Table 1: Number of people who passed the test for each level of van Hiele’s geometric thought and its passing
rate for the experimental group and the control group
With reference to Table 1, it can be found that, in the experimental group, the numbers of people who
passed level 1 had increased from 22 to 28, the numbers of people who passed level 2 had increased from 17 to
28, and the numbers of people who passed level 3 had increased from 0 to 1. It was shown that the students had
a god performance at level 1, the passing rate was quite high, up to 71% for pre-test, and for post-test, it had
reached to 90%; as to the performance at level 2, the passing rate was 55%, and for post-test, it had significantly
improved to 90%, meaning that most of students had met the criterion for level 2; as for level 3, the passing rate
was very low, the passing rate for pre-test was 0%, and for the post-test, it was 3% only.
In the aspect of the control group, the numbers of people who passed level 1 had increased from 21 to
25, the numbers of people who passed level 2 had increased from 15 to 21, and the numbers of people who
passed level 3 had increased from 0 to 3. It was shown that the students had a god performance at level 1, the
passing rate was quite high, up to 75% for pre-test, and for post-test, it had reached to 89%; as to the
performance at level 2, the passing rate was 54%, and for post-test, it had improved to 75%, although the
improvement was significant, but if comparing to the experimental group, i.e. improved from 55% to 90%, it
can be found that due to WPBL the performance of level 2 of van Hiele’s geometry test for the experimental
group was significantly superior to that of control group which had the five stages of learning model introduced
by van Hiele; as to the performance as level 3, the passing rates were very low, wherein, 0% for pre-test and
11% for post-test.
Comparing with the numbers of passers for pre-test and post-test in the experimental group and the
control group, although the numbers of passers were increasing at all levels, however, the experimental group
made significant greater improvement than the control group, in light of the above results, WPBL used in this
study indeed has big advantage to improve students’ level of the geometric thought.
Level 1 of van Hiele’s geometric thought emphasized that children were capable of distinguishing
graphics by observing the contours of specific things and the second level of van Hiele’s geometric thought
emphasized that school children would be able to distinguish the characteristics of graphics by observing the
constituent elements of the graphics and relationship between these elements. According to the reflective journal
nd
to record teaching activities, when students in the experimental group encountered a potential problem derived
from a task assigned by the executive team leader, they always could find a way of solve this problem, such as:
looking for information online, discussing problems with other team members...etc. to solve the problems; and
in the problem-solving process, they could also learn more about the graphics features, so that they might
having superior capability to explain each graphics feature by means of observation of various graphics (R0402,
R0416).
The students assigned to the control group were becoming more dependent on the teacher's examples
and answers, students seemed to expect the teachers to explain the whole problem-solving process, but when
they encountered a similar problem next time, they might be still raising hands to ask questions (R0327, R0403).
It can be found that, with reference to above results, the teaching method of PBL adopted in the experimental
group can indeed stimulate students' thinking as well as problem-solving skills, thus, it will contribute to
improve students' geometric thinking levels.
2) T-Test to evaluate the pre-test and post-test scores of experimental group and control group
According to the results of pre-test, this study conducted an independent-samples t-test to analyze
whether there were differences in the starting point for learning behavior and each level of geometric thinking
between two groups of students, the statistical analysis results were presented in Table 2. The results showed
that there were no differences in the starting point for learning behavior of two groups. After conducting the
relevant experiments, the results relating to the influence on each level of geometric thinking was shown in
Table 3. It has been found that: the performance of the second level of geometric thinking for the experimental
group was significantly superior to that of the control group; however, the differences for first level and the third
level of geometric thinking were not so significant.
Table 2: The comparison of scores for pre-test of each level of geometric thought and total scores of the
experimental group and the control group
Table 3: The comparison of scores for post-test of each level of geometric thought and total scores of the
experimental group and the control group
nd
The second level of van Hiele’s geometric thought stressed children might identify various graphics
characteristics by observing the constituent elements of the graphics and relationship between these elements. It
can be found according to the reflective journal to record teaching activities that the tasks of PBL assigned by
the teachers had usually no fixed answers, therefore, students in the experimental group could freely develop
their own creative ideas and designs, and furthermore, they even could apply the learned graphic characteristics
to their everyday lives; meanwhile in the process of specific project, they would have better understanding of
various graphic features and constituent elements and be able to utilize them more accurately (R0402, R0416,
R0507, and R0514). Besides, most of students in the control group were taught with the examples of textbooks
and exercises, although students had really learned textbooks very well, but the experimental group performed
far better than the control group in the application of knowledge (R0508, R0515). It indicates that the teaching
method of PBL adopted by the experimental group indeed can contribute to improve the students’ second level
of van Hiele’s geometric thought.
V. Conclusion
According to the results of the research, it showed that: due to WPBL the performance of level 1 and 2
of van Hiele’s geometry thought had improved significantly, wherein, the experimental group was significantly
superior to that of control group; as to the performance as level 3, the influence was very little. The Level 1 of
van Hiele’s geometric thought emphasized that children were capable of distinguishing graphics by observing
the contours of specific things and the second level of van Hiele’s geometric thought emphasized that school
children would be able to distinguish the characteristics of graphics by observing the constituent elements of the
graphics and relationship between these elements. The students in the experimental group were encouraged to
actively construct mathematical knowledge, in which, students must slowly study about the appearance and
characteristics of various graphics in the process of observation and analysis, and furthermore, they even could
apply the learned graphic characteristics to their everyday lives; additionally, when encountering the difficulties
or problems, students also must find a way to resolve this problem by themselves ; and in this way the high-level
thinking skills will be enhanced in the process of problem finding and problem solving, again and again, hence
,it is proven that the WPBL indeed can improve the school children’s level of geometric thought.
REFERENCES
[1] D. B. Wu, “Elementary school students in the graphics space takes the concept of perceptual manipulated,
mapping, theoretical description of the understanding of the follow-up study, “Research Project of the
National Science Council of the Executive Yuan, Project No.: NSC 111 92-2521-S-142-004, 2003.
nd
[2] D. Moursund, “Project-based learning using information technology,” Eugene, OR: International Society for
Technology in Education, 2002.
[3] H. Shyu, “How to use the Internet to help children become master? Network project-based learning and
teaching innovation,” Taiwan's education, 607, pp. 25-34, 2001.
[4] J. C. Bobango, “The van Hiele levels of geometric thought and student achievement in standard content and
proof writing: The effect of phase-based instruction." (Ph.D. Dissertation, The Pennsylvania State
University, 1987), UMI Publications.
[5] J. P. Becker and C. Selter, “Elementary school practices. In A. J. Bishop et al. (Eds.),” International
Handbook of Mathematics Education, pp. 511-564, Dordrecht: Kluwer, 1996.
[6] J. S. Krajcik, C. M. Czerniak and C. F. Berger, “Teaching science in elementary and middle school
classrooms: A project-based approach,” pp. 454. New York, NY: McGraw-Hill, 2003.
[7]
J. Shaughnessy, Michael and F. Burger. Willam, "Spadework Prior to Deduction in Geometry."
Mathematics Teacher, 78, pp. 419-28, September, 1985.
[8] J. W. Thomas, “A review of research on project-based learning,” 2002. Retrieved from
http://www.bie.org/index.php/site/RE/pbl_research/29
[9] L. C. Chang and G. C. Lee, “A team-teaching model for practicing project-based learning in high school:
Collaboration between computer and subject teachers. Computers & Education, 55(3), pp. 961-969, 2010.
[10] M. Basbay and A. Ates, “The reflections of student teachers on project based learning and investigating
self evaluation versus teacher evaluation,” Procedia Social and Behavioral Sciences, 1, pp. 242-247, 2009.
[11] M. L. Crowley, “The Van Hiele Model of the Development of Geometric Thought. In M. M. Lindquist &
A. P. Shulte (Eds.),” Learning and Teaching Geometry, K-12. 1987 Yearbook (pp. 1-16). Reston, VA:
National Council of Teachers of Mathematics, 1987.
[12] M. Papastergiou, “Digital Game-Based Learning in high school computer science education: Impact on
educational effectiveness and student motivation,” Computers & Education, 52(1), pp. 1-12, 2009.
[13] P. M. Van Hiele, “Structure and insight: A theory of mathematics education,” Orlando, FL: Academic
Press, 1986.
[14] R. V. Hadgraft and N. Scott, “Engineering education is problem-based or project-based learning the
answer?” Australasian Journal of Engineering Education, pp. 2-14, 2003.
[15] Usiskin, Z. P., “Van Hiele levels and achievement in secondary school geometry,”ERIC Reproduction
Service No. ED 220 288, 1982.
[16] W. Hafner and T. J. Ellis, “Project-based, asynchronous collaborative learning,” Proceedings of the 37th
Hawaii International Conference on System Sciences. Piscataway, NJ: IEEE, 2004.
[17] W. H. Kilpatrick, “The project method,” Teachers College Record, 19(4), pp. 319–335, 1918.
nd
Metadata Integration for Centralized Repository using Dspace
LertchaiKhongamnuaisak
Department of Computer Engineering
Faculty of Engineering, Chulalongkorn University
Bangkok 10330, THAILAND
[email protected]
NakornthipPrompoon
Department of Computer Engineering
Faculty of Engineering, Chulalongkorn University
Bangkok 10330, THAILAND
[email protected]
Abstract :
Nowadays, information used in any organization is usually stored in different formats and may earn
frominternal and external organization. Such information may be not available for a sudden use. Thus, it is
necessary to integrate them to create a central repository and provide a mechanism to present the results
relevant to the user need. This research presents a method to construct a centralized repository based on Dublin
Core Standard used for metadata identification and metadata structure. Also, OAI-PMH was used to import all
data from different repositories. In addition, Dspace was customized and used to establish centralized
repository. The centralized repository supports the retrieval of information efficiently to serve different needs.
Keywords-component; Metadata Integration, Centralized Repository, Library, Dublin Core Standard, Dspace
I. INTRODUCTION
Information is important because it helps to present information in a variety of research areas. But
because the data is collected by different repositories makes it not easy to find. The basic characteristics of the
data used describe or metadata. Characterized by the different name that is in the repository, such as "Author"
Some may call. "Contributor" but the same meaning so maybe misleads users. Therefore, it should be collected
or integrate data from different repository to the centralized repository, so that users can find information from
the centralized repository.
This research proposes the integration of metadata from various repositories that are stored in the
library. There pository would be designed based on the Dublin Core Standard. The integration of multiple
repositories, with the help of OAI-PMH protocol [1] and Dspace [2][3], enable an access to the centralized
repository for research, retrieve or display of the data.
nd
II. RELATED WORK
A. Digital Repository
Digital data storage resources are the main principle of the digital archive. Therefore, it is necessary to
understand the basic elements of it. Digital data repository architecture consists of three main components: user
interface, services, and digital data storage. Services are the connection between user interface and repository.
Detailed in Fig 1.
Figure 1. The architecture of a digital library [4].
The basic elements of digital data in the repository [4] should have the basic elements to identify and
describe information in digital archives, such as detailed information, description of metadata, and collaboration.
B. Dspace
Dspace [5] is an open source that has been developed by the Massachusetts Institute of Technology and
the Hewlett – Packard, which aimed to create a robust repository. The data structure of a Dspace is the best
storage stabilization. It has powerful feature such as the right to access to information, data management and
users can query data from the repository.
The research of G. Pirounakis [6] presents a comparison between different tools that are used for
storage as anarchive. This study compares features of five different repositories: Dspace, Fedora, Greenstone,
Keystone and EPrints. The result appears that Dspace is most appropriate for use than other tools. The two most
important features used in this study as 1) Metadata and digital content storage 2) Interoperability features,
which has a score that is better than the other tools. Thus, We were chosen for the preparation of a centralized
repository using Dspace. The five repositories were graded for each characteristic. The minimum score is 1 and
the maximum is 5.Table 1. In general, DSpace has almost the highest score for every characteristic.
nd
TABLE I. COMPARISON OF THE FIVE REPOSITORIES.
C. Large scale repositories
The research of large-scale repositories [7] presents a case study of the application for the integration
of data from different data sources. The methodology of the research represents used to collaborate and
Interoperability. Data are imported from various repositories by using OAI-PMH. The OAI-PMH Protocol is
responsible for harvesting and storing the data into the repository. This research is a study of the integration of
data from various repositories. The model is proposed as a model of “cooperative and separation” of the data
into the repository from various sources. Data obtained from institution A, B, and C are gathered in one place as
workflow. With the help of OAI-PMH, local repository or Dspace is created for use. See Fig. 2
Figure 2. Data flow model for large repository.
nd
D. The OAI-PMH Protocol
The OAI-PMH Protocol [8][9] can be to develop and support the interoperability between the systems.
The purpose of the OAI-PMH is a performance for connection to repository. The OAI-PMH Protocol is an open
system that can provide access to the information. They are also used for harvesting the metadata information in
the data sources are stored in the repository. In The OAI-PMH is a key part consists of two main sections.
1) Data Providers are data or metadata into the repository. It can provide the information.
2) Service providers are harvester for metadata service is provided.
E. The Dublin Core Standard
The Dublin Core standard [10] is a standard for describing electronic information. The definition of the
set of data defines the fundamental nature of the information. It is defined by the Committee Dublin core and a
group of librarians in 1995. According to the definition, the standard explains the data in an easy and
understanding way. Meanwhile, the search and retrieve are finding to complete data because this standard helps
to definition data. The Dublin Core Standard consists of the basic 15 elements as follow: 1) Title 2) Author
3)Subject 4) Description 5) Publisher 6) Contributors 7) Date 8) Type 9) Format 10) Identifier 11) Source
12)Language 13) Relation 14) Coverage and 15) Right.
III. METHODOLOGY
The main objective of this research is to propose a method for integrating metadata to create a central
repository of data sources. There are four steps: 3.1) Design standard metadata for a central repository 3.2)
Gather data from various repositories using OAI-PMH 3.3) Propose an approach to design a centralized
Repository and 3.4) Test and evaluate the collected data to a central repository. (see Fig. 3)
Figure 3. Four steps for the integration of metadata.
nd
A. Design standard metadata for a central repository
At this Step, the standards for central metadata are required for the centralized repository. In order to
establishthe standards, metadata from various repositories need to be gathered as shown in Fig 4.
Figure 4. Four steps for the integration of metadata.
After metadata
adata has been gathered, classification of each metadata is required to examine which
metadata fields are compatible with one another or has similar semantic content. For example, data from source
2 contains all of the metadata field while data from source 1 does not have date and source. Table 2 shows how
to define and design a central metadata standard.
TABLE II. FOR EXAMPLE MAPPING METADATA FIELDS COMPATIBLE EACH REPOSITORY.
The Metadata Standard is designed as reference to the Dublin Core standards. To define or identify a data set
name, metadata fields are used as references. (see Table 3)
nd
TABLE III. THE STANDARD METADATA FOR A CENTRAL REPOSITORY.
B. Gathering data from various repositories using OAI-PMH
OAI
In the process of harvesting data from various repositories to a central repository, data are imported by
amechanism OAI-PHM
PHM protocol. It is necessary to have metadata integration. The main purpose of harvesting
data in displayed Fig 5.
Figure 5. Data are
ar imported by a mechanism OAI-PHM protocol.
The mechanism of OAI-PMH
PMH can interoperate repositories, which means that the user can access data of one
repository and exchange metadata with other harvesters. The metadata base on XML format as Fig 6.
nd
Figure 6. The metadata format for repository exchange.
C. Proposes an approach to design a centralized Repository
This step proposes design for creating a central repository. The first step will be to setup Dspace [8] for
a centralized repository and data set name
name convention collections in order to customize the metadata according
to the standard metadata specification. For the storage of the data obtained from the harvested data see Table 4.
TABLE IV. FOR EXAMPLE, SETUP NAME A NEW CONVENTION COLLECTIONS.
The next step is to customize the metadata that defines the standards metadata. Dspace can be customized into
metadata as shown in Fig 7.
Figure 7. Customizable metadata in Dspace.
nd
The final step is tuning performance OAI-PMH
OAI
within Dspace. Data to the metadata
etadata of external data
stored in a centralized repository is shown in Fig 8.
Figure 8. Customizable OAI-PMH
OAI
in Dspace.
D. Testing and Evaluation of the collected data to a central repository
Data test are derived from various repositories. Researcher can search or browse in a central repository.
Theresults would display on the user interface. For Example, the number of available data is shown in Table 5.
TABLE V. FOR EXAMPLE, RESULTS THE AMOUNT OF METADATA
METADATA INTEGRATION.
The Result of metadata integration from various repositories is displayed in Dspace see Fig 9.
Figure 9. For example, results the amount of data available in Dspace.
A centralized repository of Dspace can display the search results of listed metadata. Find Fig 10.
nd
Figure 10. The search results in Dspace.
IV. FUTURE WORK AND CONCLUSIONS
The results of this research will be a centralized repository that integrates data from various repositories
if four steps are followed. Firstly, The designing of standard metadata for central repository is required.
Secondly, data need to be gathered from various repositories using OAI-PMH. Thirdly, proposes of an approach
are needed to design a centralized Repository. Finally, testing and evaluation of the collected data is required to
ensure that everything is work accordingly. Therefore, users can search and display results from a centralized
repository by metadata Integration using Dspace. In the future, it is possible that there would be further
development of integrated data in other formats such as audio and video files in which could be helpful in other
areas.
REFERENCES
[1]
Initiative,“Open
Archives
Initiative
Version
2.0:
Reference”
[Online],
Available
from:
http://www.openarchives.org, 2008.
[2] Ramachandran, “Enabling Knowledge Sharing with an Institutional Repository”, International Conference
on Advanced Learning Technologies (ICALT 2007), 2007.
[3] Dspace, Dspace Respository , http://www.dspace.org.
[4] Ribeiro-Neto, R.B.-Y.a.B., Modern Information Retrieval. 1999, Addison Wesley.
[5] Dspace, DSpace 1.7.1 System Documentation, 2011.
[6] G. Pirounakis, M.N, “Comparing Open Source Digital Library Software”, 2009.
[7] A. Mikeal, J.C, “Large-scale ETD repositories: A case study of a digital library application”, JCDL, June
15- June 19, 2009.
[8] Hochstenbach, P, “The OAI-PMH Static Repository and Static Repository”, IEEE, 2003.
[9] Society, I.C., “Search Engine Coverage of the OAI-PMH Corpus”, Mar-Apr, 2006.
[10] The Dublin Core Metadata Initiative, “Dublin Core Metadata Element Set, Version 1.1: Reference
Description,” [Online], Available from: http://dublincore.org/documents/dces/index.shtml, [June 2012].
nd
A Conceptual Model for Chinese idioms: An OWL DL Ontology for Bilingual Learners
Haitao Li Sven Hartmann
Department of Informatics
Clausthal University of Technology
Clausthal-Zellerfeld, Germany
[email protected] [email protected]
Abstract :
Motivated by the desire to solve the problem of automatically identifying Chinese idioms for English idioms and
English description, we have tested the hypothesis that conceptual models for the language patterns can
formalize the Chinese idioms domain knowledge by means of OWL DL ontology’s. Moreover, we have
represented the domain knowledge including encoding frequently used Chinese idioms and distinguishing
synonyms of Chinese idioms. According to ontology’s' standards, we have evaluated the novel OWL DL
ontology's Tbox. In addition, we encourage that the users coin new English idioms so as to benefit the bilingual
learners and researchers more by using the method of ontology’s.
Keywords-TBox; OWL; Chinese idioms; conceptual model; ontology
I. INTRODUCTION
KR (knowledge representation ) is important not only in the trade of IT (information technologies) and
Internet Technologies, but also for handing educators' knowledge down, for the deeper researches of ancient
history, and for the spread of human culture. One of the reasons is that KR can standardize basic domain
concepts and their relations for a community. Furthermore, the semantic web has became a more and more
popular research area because of the joint efforts by those including educators, researchers and participants of
industry. In addition, fruitful research results on KR (knowledge representation ) attract a lot of attention that
encourage the researchers in this field to work harder in response to public demands. As a core part of the
semantic web, ontologies are useful for organizing knowledge, sharing knowledge, retrieving information, and
managing knowledge. In short, enhancing ontologies can improve the research on KR and the technologies of
ontologies have the ability to shoulder the task of identifying concepts.
Facing this trend, we try to help not only English-Chinese bilingual learners, but also the Western
English speakers who would like to identify CIs (Chinese idioms) for EIs (English idioms) description, to
benefit from a well-designed OWL DL ontology. Moreover, Chinese idioms ( the language or dialect of Chinese
people, region ,class, etc. ) are considered as the cultural heritage and essence of China. Additionally, it is a
global communication tool to use English for the world trade and economy, that is why our model is based on
the English language. We propose a new ontology that CIs' concepts are represented in a clear, modelled,
machine readable way. By this effort, the hypothesis of the possibility that conceptual models work in Chinese
idioms domain can be explored by a substantial amount of work.
We are trying to seek the common ground between CIs' concepts and EIs' descriptions, though they
come from different cultural backgrounds. It is not strange that there are a lot of CIs' concepts missing
corresponding EIs' descriptions, mainly due to the cultural reasons. Nevertheless, we have obtained the common
nd
ground on the issue of identifying CIs' concepts for EIs description while leaving aside the differences
mentioned above.
As a first contribution, our research has created an OWL DL ontology which adds to knowledge in a
way that links Chinese idioms to English idioms.
As a second contribution, from a conceptual modelling view this research has provided a description
logic knowledge base that consists of both a TBox and an ABox.
As a last contribution, we have added new knowledge to KR ( knowledge representation ) area that will
promote the international culture interchange.
II. RELATED WORK
We have summed up not only the related researches but also the relevant software systems.
A. Two types of researches
With regard to the related work on automated identification of idioms and / or the relevant semantic
recognition researches, we have summarized two types of researches according to the difference of their
emphases that are predicates and / or definitions.
As the first type of the researches, it is the predicates-centered researches. They have formed our
research's theoretical bases.
One of the interesting papers has shown that the integration of concepts is a classification and / task
[1]. Similarly, “An analysis that defines predicates for Wordnet verb classes and link them to semantic
interpretation is presented.” [2] The idea is that “the meaning of the verb depends not only on its selection
restrictions, but also on the syntactic relations that realize them”[2]. “For every verb in a clause, Word net
provides a list of verb synsets for which we have defined predicates.”[2] “The predicate that realizes the most
syntactic relations in the sentence is selected as the meaning of the verb.”[2] As one of the research results ,forty
eight frequently used verbs was presented, moreover, “a model for semantic interpretation that integrates the
determination of the meaning of verbs, the attachment and meaning of prepositions, and the determination of the
thematic roles” was shown in [3]. Facing a common debate, the authors respond like this: “A critique that can be
levelled against our approach could be that it needs the hand-crafted construction of verb predicates, which is a
rather difficult and time-consuming job. The reply to this is that once the verb predicates are defined, they are
defined for every natural language application. This is so because their definitions are not tied to any given
application, and their selection restrictions are based on a general ontology of English. Moreover, as we have
indicated in this paper the semantic interpreter can be used for automatically acquiring inference rules to be used
by the knowledge acquisition system.”[4] With respect to the testing, the authors asserted that “we have
conducted diverse tests and the system has been demoed in a major conference.”[5] At the same time, “The
algorithms were tested to determine how successful they were in disambiguating the nominalization, and in
recognizing the underlying verbal concept of the nominalization and filling its thematic roles.”[6]
As the second type of the researches, it is the definitions-centered researches. They link up with our
research in the way that they have provided definite clues of distinguishing Chinese idioms and / or their
synonyms from each other.
nd
An interesting research has tested that “the extended entity-relationship modelling can provide both a
well defined semantics and improved modeling elements that naturally reflect English language sentence
structures.”[7] In other words, the users require more than three different types of information, the requirement
'hierarchy can be represented by higher-order object types or a directed graph. For our research, the implications
of this work is the following: The description of the target database “is usually given in some natural language
,for instance in English. Conceptual design aims at transforming English sentences into a conceptual database
schema. A conceptual data model should therefore possess modeling features that can represent English
sentence structures.”[7] In order to meet the requirements of intelligent question-answering, an important article
has triedto resolve the problem of various domains' knowledge representation in the form of ontologies [8]. The
authors have set up an application scenario for mobile users that the ontologies are relevant to information
services on the web. In the end the authors have concluded that it is of great importance to keep conceptual
clarity .For our research's precise measure of the semantic similarity, we have analyzed the literature on the
algorithms of information matching in [9], [10], [11], and [12].
B. Four kinds of software systems ( From the perspective of idioms, we have conducted a brief classification of
the existing search engine tools on the web that are freely and readily accessible. )
•
The major search engine such as Google as the first class of idioms identification tools.
For fixed idioms and / or phrases, Google's strategies include that providing different sets of the latest
example with and without quotation marks to search for idioms. But for Chinese idioms, the
translate.google.com website only shows the users the corresponding Pinyin.
•
Some online websites like the www.phrases.netand the www.makeuseof.com/dir/idiomsite-meaningsofcommon- sayingsas the second class of idioms identification tools.
They provide the service that allow the users to find out the meaning of a particular idiom and / or
phrase.But they do not serve as an idiom matcher or finder to identify a Chinese idiom. Furthermore,
their qualities of service is poor, at the same time, their functions are incomplete.
•
The retrieval systems such as OCIRS, Shunfeng, Wuyou, etc. as the third class of idioms identification
tools.
The authors of OCIRS claim that the system can search idioms via semantics. While we have tried and
noticed that no definite retrieval results are shown.
•
Others. ( Including the www.talkiteasy.org, the englishpond.com , the ask.metafilter.com , the
www.usingenglish.com,
the
www.phrases.org.uk
,
the
www.thefreedictionary.com,
the
idioms.thefreedictionary.com , the englishteacherfred.com etc. )
All of them are connected to idioms, but none of them can identify Chinese idioms for English idioms and / or
English description as far as we know.
III. THE OWL DL ONTOLOGY
The goal of this OWL DL ontology is to capture knowledge in the Chinese idioms domain. Our
ontology describes the concepts of 1000 frequently used Chinese idioms and also the synonymous relationships
that hold between those concepts. By using Protégé OWL, we define and describe those concepts so that we can
distinguish them from each other for English idioms and descriptions. Furthermore, the logical model allows
nd
that we can use the related tools to recognize which Chinese idioms fit under which definitions that are the
combinations of Longman defining words. In addition, employed in our design as a sidekick, Wordnet
contributes its English words' synsets for the mapping between English descriptions and Chinese idioms'
definitions. The reason for using Wordnet is that our ontology 's atomic element is the English word and that the
similarity between English words are computed based on Wordnet' ssynsets.
The novelty of this ontology is that it allows new English idioms and new Chinese idioms to be coined
if the new idiom is articulated. The first reason for that our ontology can accept and append new English and / or
Chinese idioms is the following: we spare categories for possible new idioms based on Wordnet'ssynset, in other
words, each word in our ontology has its own category and / or variable set preserved, and the category has a
oneto -one mapping to the words in Wordnet' synset. The second reason is that each possible synonym has its
own category in our ontology so that the slightest difference between the existing idioms and the foreseeable
new idioms can be captured. Indeed, it is one of our ontology's features that the flexibility is realized through
spared categories.
The OWL DL ontology is designed to contain 12000 classes, 10 properties, 1000 individuals, and
2014axioms. With regard to the 12000 classes, 1000 of them constitute the classes of Chinese idioms; 3000 of
them comprise the classes of the synonyms' distinctions; 4000 of them make up the set of the defining words;
and the last 4000 of them compose the set of the defining words' merged synsets ( in accordance to the core of
Wordnet ). With respect to the 10 properties, they are the following relations between the classes: hasAD;
hasEID;hasDescription; distinguished-from; encodes; encoded-by; composes; is-composed-of; part-of; and haspart. Among them, the last six properties has a relationship of “inverses”; the last two properties' ( part-of and
haspart) superproperties are composes and is-composed-of respectively. With respect to the 2014 axioms, 2000
of them are the “disjoint” relationship between the defining words' classes; 14 of them are the “covered by”
connection to their discrete restrictions. The reason for choosing OWL DL ( description logic ) is that we
determined to trade the property of fine-grained ontologies for the decidable inferencing.
With respect to our ontology's storage, our design rationale trade space for time. The description logic
reasoned infers sub assumption relations on the Tbox. Additionally, rule logic capture the assertions on the
Abox. In this way, both inferred Chinese idioms' definitions and the original assertions of Chinese idioms can
be materialized in relational database. Consequently, this kind of design can decrease the response time because
the data are ready after the loading of this ontology.
IV. MATERIALS AND METHODS
The fundamental idea for the English descriptive ontology is the following: First, we encode each
Chinese idioms into a four-word group (a verb, a subject, an object, and an adjective) consisting of the
Longman defining vocabulary, which has been carefully selected. Second, we organize both the Chinese idioms
and the defining words into a tree structure in Protégé-OWL. Third, for the sake of distinguishing the synonyms
of Chinese idioms, we consider to add one more layer in the sub trees for the following reasons: each member in
the Chinese idiom's synset (always they appear in pairs) can be differentiated from each other by the style, the
attitude, and the degree. Furthermore, they have rarely the same sense in all the above ways. For example,
hun2bu4shou3she4 and ru2zuo4zhen1zhan1 ( the numbers represent Chinese language's four tones ) are a pair
of synonyms, the former's closest corresponding English idiom is that “to have ants in one's pants”, while the
nd
latter's, “like ants on a hot pan”. We can understand that the former refers to “the spirit leaves the body”, and the
latter refers to “feel extreme anxiety”. We analyze from the ancient Chinese standpoint as follows: their
emphasizing points are different, the degree of the latter is stronger, furthermore, from the attitude point of view,
people disapprove of the former, but for the latter no disapproving. These are the causes why we can recognize
between the synonyms which is which. Fourth, we decide to ask the user for a subject and an object. In fact, we
hesitated before the decision for the following reasons: on the one hand, as the Chinese idioms usually have no
person, no tense, and no plural, as a result, the idioms always sacrifice completeness for function; on the other
hand, the whole sense of a Chinese idiom is usually centered around the predicate, furthermore, without an
object we would not know the goal, and / or the aim, and / or the target of the action. After these consideration,
we think that the English is rule based, therefore we decide to ask for two nouns ( a subject and an object
respectively ) as the articulating of the input English idiom. For solving the ambiguity problem in Chinese
idioms, we allow the users do the following selections: Pin down the attitude, the style, and the degree
differences: A. People disapprove of this one. B. No disapproving attitude. C. In a special situation such as
literary. D. In an ordinary situation. E. A stronger one. F. No stronger degree. Note that the pairs of (A,B) and /
or (C,D) and / or (E,F) are mutual
exclusive.
The idea is a combination of a verb, a subject, an object and an adjective that is organized
systematically in aTbox of the ontology for each Chinese Idiom. In this way, to reach the goal of identifying a
sole Chinese idiom, you simply walk down the tree with at least four layers. In the first layer, it is the root of
ideas. In the second layer, the defining words of verbs in the Longman defining vocabulary are arranged as the
first step for deriving a Chinese idiom. In the third layer, we combine the previous verb and a subject. In the
fourth layer, it is the combination of the previous verb, the previous subject and an object that play the role of
the expressing a Chinese idiom's sense, the reason is that a Chinese idiom is often centered around the predicate,
and the reason for that only two nouns will suffice is the following: the particular action in a Chinese idiom is
described by the verb in the second layer, and these nouns are often the cause and the result to express an idea.
In the fifth layer, it is the adjectives' turn which always can determine the degrees of the previous action and / or
results. More strict scholars may ask that why no adverb? The reasons for this phenomenon is the following: the
underlying reason isthat an adjective can transmit the information of an adverb in ancient Chinese language,
furthermore, for most of the Chinese idioms there is no person, even no tense in the concise way of expression.
Therefore, we can achieve a better understanding about a course-grained language from the standpoint of the
West, in other words, this old language prefers to sacrifice completeness for function. Similarly, we can expect
that the users articulate theinputted idioms by viewing a pronoun as a noun, viewing an adverb or a numeral as
an adjective, and ignoring the conjunctions and the prepositions.
The definitions of the Chinese idioms were obtained from [13]. And the building tool and materials for
our ontology were obtained from the website of http://wordnet.princeton.edu/wordnet/download/standoff/ [14].
V. RESULTS
The whole picture of this research is to identify a sole Chinese idiom as a counterpart of a particular
user's inputted English idiom and description. In order to achieve the overall goal we decompose it into three
sub goals:
nd
(1) a conceptual model for language patterns, (2) a novel OWL DL ontology, and (3) the logic for identifying
the counterpart.
Figure 1. Our example of the encoding of the idioms “ru2zuo4zhen1zhan1” and “hun2bu4shou3she4”(the
numbers: Chinese language 'stones)
Figure 2. Use case diagram
Fig. 1 shows the result of a pair of Chinese idiom' synonyms, “hun2bu4shou3she4” and
“ru2zuo4zhen1zhan1” respectively. Additionally, the class “Attitude” in the picture ( on the left side and in the
nd
middle of the list of the classes ) summarized the steps of the decoding. The layers of classes shows the state of
the
post-decoding:
for
“hun2bu4shou3she4”,
it
is
composed
of
the
“Disapproving_
Attitude”,
“D_Special_Situations”, “DS_No_Stronger_Degree”, “V9”, “N1_9”, “N2_9”, and “A_9” categories
respectively;
for
“ru2zuo4zhen1zhan1”,
it
is
composed
of
the
“No_Disapproving_Attitude”,
“N_Special_Situations”, “NS_Stronger_Degree”, “V12”, “N1_12”, “N2_12”, and “A_12” categories
respectively.
We believe that logic and ontologies can shoulder the task for the use of the bilingual learners. Fig. 2,
Fig. 3,Fig. 4, Fig. 5, and Fig. 6 show the logic of the CMCI ( conceptual model for Chinese idioms). Fig. 3
shows the logic of computing for the similarity degrees. The variables stand for two English words in the same
word class,and the function E(i) stands for their synsets in Wordnet. In the equation, we compute the degree of
similarity by using the overlap between their synsets.
Figure 1.
The logic for measuring the similarity between synsets
Figure 4. The logic for determining the Chinese counterpart “ru2zuo4zhen1zhan1” (the numbers: Chinese
language's tones)
In other words, if no overlap between the two defining words' synsets, the similarity degree is zero; if
the overlap is exactly the synsets' union, the two synsets are definitely same; in other cases, the degree is a
decimal between zero and one.
Fig. 4 shows the matched conditions between users' inputted information and the ontology's definitions.
In this equation, if one of x, y, and z is false, the Chinese idioms' synonyms can be distinguished; at the same
time, if none of a, b, c, and d is false, the user's inputted English idiom and description can be matched. The
reasons for setting up the logic for determining the Chinese counterpart are to distinguish the synonyms of
Chinese idioms so that every concept is clear and definite. Fig. 5. and Fig. 6 show the algorithm of identifying
Chinese idioms.
nd
Figure 2.
The algorithm of identifying Chinese idioms
Figure 6. Our example of the matched Sample1 (Chinese idiom counterpart “ru2zuo4zhen1zhan1” ) depicted as UML object
diagram
VI. DISCUSSION
We show here that an OWL DL ontology conduces to identifying Chinese idioms for English idioms
and description. In addition, a conceptual model for Chinese idioms that allows new English idioms and / or
new Chinese idioms to be coined was the main outcome of this research. In contrast, the present state of
knowledge relating to Chinese idioms' identification is far from satisfying. For example, OCIRS's atomic
nd
element was the Chinese characters, even worse, the results were not unique [15]. In compliance with the
evaluation standards of ontologies, with regard to licensing, the CMCI ( conceptual model for Chinese idioms )
uses GNU license to promote interoperability; with respect to maturity, the CMCI will be kept on being
developed and maintained; with respect to granularity, the CMCI is a coarse-grained ontology that is based on
description logic; with respect to security, the CMCI respects users' right to privacy. Why should the
components that construct the Chinese idioms ontology have less than ten tiers? On the one hand, we think that
form follows function. On the other hand, we designed such a multi-layered structure for the convenience of
determining to which ontological category a user's inputted concept has to belong.
What is the nature of ontologies in the semantic web context? Ontologies exist for the purposes of
mapping
human languages to their corresponding senses. To achieve this, we have to keep conceptual clarity. Future
work will be directed to applying this novel ontology in a business context. What is the use of the CMCI in a
business context? The retrieval of the Chinese idiom counterpart is the main function for the use of the bilingual
learners. In the educational sector, a conceptual model with the following properties is required. The principle of
the conceptual model is to solely and definitely encode the frequently used Chinese idioms, so that we can
ensure the clarity of each concept. The purpose of the conceptual model is to encode frequently used Chinese
idioms so that we can definitely identify them by using the combination of the Longman defining words. In
addition, the reason for building an OWL DL ontology is to represent the Chinese idioms' domain knowledge
and share the resources with anyone who is interested in the semantic web.
ACKNOWLEDGMENT
Thank my original country and university for supporting me sustainably, anyone for kindness to our research.
REFERENCES
[1] F. Gomez, and C. Segami, “The recognition and classification of concepts in understanding scientific texts,”
Journal of Experimental and Theoretical Artificial Intelligence, vol. 1, pp. 51–77, 1989.
[2] F. Gomez, “Linking WordNet verb classes to semantic interpretation,” Proceedings of the COLING-ACL
Workshop on the Usage of WordNet on NLP Systems, Universite de Montreal, Quebec, Canada, pp. 58–64,
August 1998.
[3] F. Gomez, C. Segami, and R. Hull, “Determining prepositional attachment, prepositional meaning, verb
meaning and thematic roles,” Computational Intelligence, vol. 13(1), pp. 1–31, February 1997.
[4] F. Gomez, and C. Segami, “Semantic interpretation and knowledge extraction,” Knowledge-Based Systems,
vol. 20, pp. 51–60, 2007.
[5] F. Gomez, “An algorithm for aspects of semantic interpretation using an enhanced WordNet,” Proceedings
of the 2nd Meeting of the North American Chapter of the Association for Computational Linguistics,
NAACL-2001, CMU, Pittsburgh, pp. 1–8, 2001.
nd
[6] R. Hull, and F. Gomez, “Semantic interpretation of nominalizations,” Proceedings of the Thirteenth National
Conference on Artificial Intelligence, Portland, Oregon, pp. 1062–1068, August 1996.
[7] S. Hartmann, S. Link, “English sentence structures and EER modelling,” In Conceptual Modelling APCCM 2007, Australian Computer Society, CRPIT, vol. 67, pp. 27-35, 2007.
[8] D. Oberle, A. Ankolekar, P. Hitzler, P. Cimiano, C. Schmidt, and M. Weiten, etc., “DOLCE ergo SUMO: on
foundational and domain models in SWIntO (SmartWeb Integrated Ontology),” Technical Report, AIFB,
University of Karlsruhe, pp.1-41, 2006.
[9] D. Beeferman, A. Berger, J. Lafferty, “Statistical Models for Text Segmentation,” Machine Learning Journal
( Special Issue on Natural Language Learning), vol. 34, pp. 177-210, 1999.
[10] D. Beeferman, A. Berger, J. Lafferty, “A lightweight punctuation annotation system for speech,” In
ICASSP 1998, pp. 689-692, 1998.
[11] D. Beeferman, A. Berger, “Agglomerative clustering of search engine query log,” In KDD 2000, pp. 1-10,
2000.
[12] D. Beeferman, “Lexical discovery with an enriched semantic network,” In ACL/COLING 1998,
Proceedings of the Workshop on Applications of WordNet in Natural Language Processing Systems, pp.
135-141, 1998.
[13] H. Xu, Little Chinese-English dictionary. Beijing: Foreign language teaching and research press, April,
2004.
[14] G. A. Miller, “WordNet: a lexical database for English,” Communications of the ACM, vol. 38(11), pp. 3941,1995.
[15] H. B. Hu, C. M. Tu, C. L. Fu, L. Fu, F. Mao, & Y. Ma, “OCIRS: An Ontology-based Chinese Idioms
Retrieval System,” World Academy of Science, Engineering and Technology, vol. 42, pp. 641-647, 2010.
nd
The Impact of Technology Enhanced Learning in Southwestern China:ca Case Study
Haitao Li Sven Hartmann NielsPinkwart
Department of Informatics
Clausthal University of Technology
Clausthal-Zellerfeld, Germany
[email protected]
Abstract :
Technology-enhanced learning is increasingly used in developing countries, but is often not systematically
evaluated. This paper presents a case study at Sichuan International Studies University where an
“eExamsystem”, a self-training and test system, has been used. Results show that the eExam system helped
poorer students from rural areas catch up with their peers from urban areas.
Keywords-eExam; eLearning; technology-enhanced learning; evaluation
I. INTRODUCTION
Technology-enhanced learning is increasingly being adopted in developing countries, often in order to
offer children also from poorer regions a chance to learn even if resources for teachers and schools are scarce.
With regard to the definition of developing countries, the statement “most countries in Latin America, Middle
East, Africa, Southeast Asia, and some parts of southern Europe, are designated as ‘developing countries’
because of their lower rank in the United Nations Development Program (UNDP) Human Development Index.”
can be found in [2].
In China, higher education has been established throughout a period of at least one hundred years.
Additionally, about five hundred universities located in more than fifty cities have trained a large number of
experts / professionals in many scientific disciplines. However, a gap exists between these achievements and, at
the other side, the fact that chances of urban school aged children going into higher education are three or four
times greater than for children from rural areas. According to [2], China has 70,000 schools with computers and
more than 10 million students who have mastered basic computer skills, but most of these schools are in cities,
not in poorer, rural areas.
Currently, it is necessary to introduce technology-enhanced learning in such a rapidly developing
country as China. The main reason is that the universities are undermanned, but there are also other reasons for
the necessity to use technology-enhanced learning in developing countries: (1) the development of learners'
capabilities and their future careers require information technology. (2) the Chinese national teaching
programme and quality oriented education include information technology. (3) information technology can lead
to shorter exam times. When freshman numbers increase year by year, the students would hardly feel
convenience without the support of information technology. Furthermore, it is feasible that information
technology could foster the learning of unprivileged freshmen for the following reasons: (1) learning
information technology is especially prevalent among younger teenagers. (2) Chinese National Computer Rank
Examinations have successfully been implemented every year since the 1990s. (3) the existing eExam products
developed by the Ministry of Education of China can measure achievements in a variety of different fields,
nd
including office applications, VB programming, VFP programming, VC++ programming, network applications
etc.
In this paper, we describe the above-mentioned eExam system, which is designed to provide equal
educational opportunities for freshmen students from the rural areas in China.
Our aim is to validate the eLearning effect for the lower income regions of China - we therefore
investigated the association between class grades and the area the students came from (low/high income). The
data came from practical eExams about first-semester office application courses at Sichuan International Studies
University. We put forward the hypothesis that the eExam system does enhance the learning of those freshmen
who come from underdeveloped regions. In other words, we think that the eExam system plays a positive role
for rural freshmen students, especially when the number of students is increasing and the universities are
undermanned.
About the organization of the next sections, first, we are going to describe some related work in other
developing countries. Second, we would like to briefly introduce the eExam system. Third, we will present the
data collected from two year’s teaching practice. Fourth, we will argue that having using the eExam system has
made all the difference for poorer students from rural areas. Fifth, we would discuss our outcomes, our
limitations, and the potential importance of this paper.
II. RELATED WORK
In order to understand the current eLearning growth points in developing countries, we have conducted a
literature search using the ISI Web of Knowledge and scholar.google.com. While a recent, larger scale survey
does not seem to exist, some papers attempted to suggest what kind of functions should be provided to raise the
online learning satisfaction of underprivileged students [3], [4], [5], [6], and [7]. Nevertheless, these papers did
not deal with the validation of eExam tools or skills about office applications. From Table I we can see that most
eLearning systems are for one particular skill of the trainees. Furthermore, a quite high proportion of the papers
apparently took it for granted that learning technology would necessarily result in desired effects [8], [9],
[10],[11], [12], [13], [14], [15], and [16]. On the contrary, some papers doubt eLearning. For example, a review
of 150 distance education programs in sub-Saharan Africa has concluded that traditional, paper-based means of
distance learning continue to be more reliable, sustainable, and widely used than online methods of learning
[17]. India’s technology and distance learning institutions, including the Indira Ghandi National Open
University (IGNOU),the Indian Institute of Technology, the Birla Institute of Technology, and the School of
Education Technology at Jadavpur University, have led the way in online degree courses and modules [1].
However, they fail to consider the need to provide basic literacy and computer skills and access to computers to
poorer groups before they can begin to engage in higher online learning. As stated in [2] on page 51, this is
problematic: “Despite the availability of technologies in urban regions, rural and poor people continue to be
deprived of investment ,infrastructure, and skilled teachers. Thus far, the introduction of computers into
education in developing countries seems to have done little to widen educational access to the rural poor, who
also do not have easy access to conventional schools and colleges.”
What happened in other Asian developing countries?
nd
In Philippine, The University of the Philippines’ Open University launched am-learning program in
2004 that offered print and mobile materials for simple and universally relevant health, literacy, and numeracy
education [18]
.A survey of 387 students in their final undergraduate year at the Virtual University of Pakistan
concluded that the majority of students found learning over the Internet and via satellite TV beneficial. The
majority also reported difficulties in accessing computers, libraries, learning environments and efficient mentors
[19].
At Indonesia’s Open University, the University of Terbuka, 320,000 students reported limitations in
online infrastructure and access. Distance education systems and students continue to rely heavily on post,
courier services, and telephones [20].
TABLE I. RELATED WORK IN OTHER DEVELOPING COUNTRIES
nd
Figure 1. The eExam system in classroom use
This paper attempts to investigate the effects of providing learning technologies to poorer students
from rural areas based on evidence collected in a large-scale
larg
field study over a 2-year
year period (see Fig. 1 for an
impression of how the class was conducted).
III. THE EEXAM SYSTEM
The eExam system, mentioned before in this paper, it is a web based learning and eAssessment tool for
the Chinese National Computerr Rank Examinations. It has gone through different cycles of iterative design
,development, and testing since the 1990s. At first, it supported only a few fields such as Fox base which was
running on MS-DOS.
DOS. As of 2007, a network version that covers many application fields exists. The
eExamsystem contains approximately 5000 examination questions and it can support as many as 2000 parallel
users. An automatic marking function characterizes one of the main features of the eExam system. This function
makes it possible for a few examiners to examine one or two thousand examinees in a few hours. If the
freshmen students wanted to do some exercise by themselves, a pool of test questions
questions for trainees are provided.
Figure 2. The eExam User’s Interface for trainees
nd
Fig. 2 shows the user interface of the system. The interface shows, among other things, the time
remaining, the questions remaining, the questions completed, etc.
Technically, the eExam system is based on a three-layer
three layer architecture. In the client layer, information
services can be requested by a browser. In such a way, end users such as students do not have to install software
on their machines. In the middle layer, application servers for various domains are packaged. The application
servers’ interfaces with uniform format are open and exposed. Thus, the complexity of operating systems and
evolutionary platform can be ignored by developers, who can be concerned with business itself. Functional
component-ware
ware has been used to reduce early construction costs and maintenance costs. All applications are
compliant with protocols such as SOAP, WSDL, and UDDI. In the hibernate layer, Object-Relational
Object
mappings
were adopted.
IV. MATERIALS AND METHODS
At Sichuan International Studies University in China, an “office application” required course was given
to all University freshmen twice a week for half a year. All the teaching materials for this course were contained
in the eExam
am system. Usually, one teacher was responsible for up to sixty students. Such a proportion meant that
only instructional technology could ensure personalized learning. As such, during the whole learning process,
only the eExam system was used in classes, and teachers merely demonstrated the examination questions in the
database of the eExam system.
When the learners began to do exercises in the classes, the system provided a set of tasks or questions
that was randomly chosen from the servers. Furthermore, a continuous assessment allowed the learners to know
the real state of their mastering skills. It was this real-world
real world course that generated all the data in this paper.
A. The First Experiment
TABLE II. DISTRIBUTIONS OF STUDENTS IN YEAR 2006/07
nd
Figure 3. Results of the class of 2006 in September of 2006
In September of 2006, ten classes of randomly chosen freshmen students (total 198 students: 89 rural
students; 109 urban students) were required to take the eExam before the course started. This was in order to test
how much office application skills that they had learned before admission. Usually, most students partly met the
elementary requirements of the course such as typing.
In January of 2007, all freshmen students (total 1635 students: 696 rural students; 939 urban students)
were required to take the same eExam. This time, most students finished all the tasks in the eExam: typing,
word processing, spreadsheet calculations, and web page construction et al. The data that whether a student
came from urban or rural areas is obtained from the office of student administration at Sichuan International
Studies University in China. Before using the electronic system, Table II. and Fig. 3 have shown that the rural
group performed not good, please notice the cross marks of the rural group and the urban group in Fig. 3 that
shows the proportions of the students' number of fail. In contrast, after using this electronic system Table II. and
Fig. 4 have shown that the proportions of the students' number of fail in rural group have dramatically
decreased. At the same time, the proportions of the students' number of very good in rural group have
significantly risen.
Figure 4. Results of the class of 2006 in January of 2007
nd
B. The Second Experiment
In September of 2007 (total 256 students: 96 rural students; 160 urban students) and in January of 2008
(total 2099 students: 782 rural students; 1317 urban students), for the new year’s freshmen students in grade
2007, the same experiment was repeated in order to validate the results.
V. RESULTS
To investigate whether the eExam system could help the poorer students’ learning of office application
skills, we analyzed the data obtained in 2006, 2007, and 2008
.A. Results of the class of 2006
Table II. and Fig. 3 show how the students from urban and rural areas performed in the pre-test and in
the post-test. The results are classified into four categories: Very Good (score 80% or higher). Good (70%79%),Pass (60%-69%), and Fail. .The data shows that no single rural student got a mark scored “Very Good” on
the pre-test. In contrast, 60% urban students received this mark. Moreover, 60% rural students failed to receive
the minimum score for passing the test (60%); in contrast, every urban student passed
.Table II. and Fig. 4 indicate that, after the class using the eExam system, the results of the rural
students were so well proportioned that they had caught up with their peers from urban areas.
B. Results of the class of 2007
We wanted to test that the similar thing would happen again in the next year. Accordingly, we tested
the freshmen students in 2007/08.
TABLE III. DISTRIBUTIONS OF STUDENTS IN YEAR 2007/08
nd
Figure 5. Results of the class of 2007 in September of 2007
Table III., Fig.5, and Fig. 6 show the similar results. These confirm the findings from the year before:
again ,the students from the rural, poorer areas were not successful at the pre-test (no student better than 70%,
45 out of 96 failing) while the students from the urban areas did much better (127 out of 160 with a “very good”,
only one student failing). At the post-test, such a difference could not be observed anymore.
In summary, this data is evidence that the class (in which the eExam system was the main learning tool,
which was also available for students to train themselves from home) did play a positive role for the
development of rural freshmen students’ office application skills.
VI. DISCUSSION
The main outcome of the work is the following. In September of 2006, the results of the first
experiment indicated an association between the scores and the areas the students came from, however, about
half a year later such association disappeared. These results could be repeated in the next year. We would have
liked to compare these results to a situation where the eExam system was not used (but traditional learning and
teaching methods) but did not have access to earlier data before 2006.
Figure 1.
Results of the class of 2007 in January of 2008
In order to improve educational equity, technology-enhanced learning tools can bridge the gap between
the rural students and the urban students. In China, the economy is rising, the number of freshmen students is
increasing, and the universities are chronically undermanned. In this situation, learning technologies, such as the
eExam system presented in this paper, are needed to help poorer students – our data shows that good computer
based learning tools can indeed be successful in this regard.
nd
With regard to the potential importance of our results, other special fields of study such as biology
could be enhanced by using on-line learning tools. For example, the sites of NCBI (the National Center for
Biotechnology Information, which advances science and health by providing access to biomedical and genomic
information),and EMBL (European Molecular Biology Laboratory) could play an important role in helping rural
university students.
VII. CONCLUSION
Overall, the field study results have confirmed our initial hypothesis that information technology can
improve the poorer freshmen students’ office application learning. In addition, this paper provides an insight
into the causes of the necessity for technology-enhanced learning at Sichuan International Studies University in
south western China.
In south western China as in many developing countries, the development of education lags behind
that of economy. It is impossible to build up a famous university in a short time. There are too many students for
the capacity of the existing universities. Therefore, using advanced learning technologies is necessary. The
paper concludes that the poorer freshmen students can achieve educational equality in computer skills by using
one such advanced learning technology: the eExam system.
Based on our results, we believe that in many western China cases, where there are conventional
colleges, the described system is workable. This conclusion does not attempt to suggest that the use of eExam
/eLearning systems should be limited to colleges or to office applications. The implication is the opposite. The
use of the system could be expanded at most secondary specialized schools in western China so that more rural
students can keep pace with the progress of the information era and achieve fulfilled lives in the future.
ACKNOWLEDGMENT
Thank my original country and university for supporting me sustainably, and Dr. rer. nat. Nguyen-Thinh Le.
REFERENCES
[1] P. Sharma, The distance in education and online technologies in India, A. A. Carr-Chellman ed., London:
Sage Publications, pp.52–66, 2005.
[2] S. Gulati, “Technology-Enhanced Learning in Developing Nations: A review,” International Review of
Research in Open and Distance Learning, vol. 9(1), pp. 43-58, Feb., 2008.
[3] C. R. Wright, G. Dhanarajan, S. A. Reju, “Recurring Issues Encountered by Distance Educators in
Developing and Emerging Nations,” International Review of Research in Open and Distance Learning, vol.
10(1), pp. 127-152, Feb., 2009.
[4] I. Jung, T. M. Wong, C. Li, S. Baigaltugs, T. Belawati, “Quality assurance in Asian distance education:
Diverse approaches and common culture,” International Review of Research in Open and Distance
Learning, vol. 12(6), pp. 63-83, Oct., 2011.
[5] A. S. Andersson, Å. Grönlund, “A Conceptual Framework for E-Learning in Developing Countries: A
Critical Review of Research Challenges,” The Electronic Journal on Information Systems in Developing
Countries, vol. 38, pp. 117-133, Sep., 2009.
nd
[6] J. Kim, W. Lee, “Assistance and possibilities: analysis of learning-related factors affecting the online
learning satisfaction of underprivileged students,” Computers & Education, vol. 57(4), pp. 2395-2405, Dec.,
2011.
[7] A. Andersson, “Seven major challenges for e-learning in developing countries: Case study eBIT, Sri Lanka,”
International Journal of Education and Development Using Information and Communication Technology,
vol. 4(3), pp. 45-62, Oct., 2008.
[8] R. Garrote, T. Pettersson, M. Christie, “LiveUSB Mediated Education: A method to facilitate computer
supported education,” Australasian Journal of Educational Technology, vol. 27(4), pp.619-632, Aug., 2011.
[9] A. S. Blignaut, J. E. Hinnostraza, C.J. Els, M. Brun, “ICT in education policy and practice in developing
countries: South Africa and Chile compared through SITES 2006,” Computers & Education, vol. 55(4),
pp.1552-1563, Sep., 2010.
[10] R. Rubio, S. Martin, S. Moran, “Collaborative Web Learning Tools: Wikos and Blogs,”
ComputerApplications in Engineering Education, vol. 18(3), pp. 502-511, Sep., 2010.
[11] R. Vyas, S. Albright, D. Walker, A. Zachariah, M. Y. Lee, “Clinical training at remote sites using mobile
technology: an India-USA partnership,” Distance Education, vol. 31(2), pp. 211-226, Apr., 2010.
[12] M. Hodgins, B. Battel-Kirk, A. G Asgeirsdottir, “Building capacity in workplace health promotion: the case
of the Healthy Together elearning project,” Global Health Promotion, vol. 17(1), pp. 60-68,Jan., 2010.
[13] F. Helali, “Using ergonomics checkpoints to support a participatory ergonomics intervention in an
industrially developing country
(IDC)--a case study,” International Journal of Occupational Safety and Ergonomics, vol. 15(3), pp. 325-337,
Sep., 2009.
[14] T. Obura, W. E. Brant, F. Miller, I. J. Parboosingh, “Participating in a Community of Learners enhances
resident perceptions of learning in an e-mentoring program: proof of concept,” BMC MEDICAL
EDUCATION, vol. 11(3), pp. 125-129, Jan., 2011.
[15] S. Raturi, R. Hogan, K. H. Thaman, “Learners' access to tools and experience with technology at the
University of the South Pacific: Readiness for e-learning,” Australasian Journal of Educational Technology,
vol. 27(3), pp. 411-427, Sep., 2011.
[16] C. M. Guerra, M. P. Ramos, V. Z. Penna, J. M. Goto, L. Q. Santi, V. A. Stempliuk, J. Sallas, E. A. Servolo
Medeiros, “How to educate health care professionals in developing countries? A Brazilian experience,”
American Journal of Infection Control, vol. 38(6), pp. 491- 493, Aug., 2010.
[17] J. Leary, Z. L. Berge, “Trends and challenges of e-learning in national and international agricultural
development,” International Journal of Education and Development using Information and Communication
Technology, vol. 2(2), pp. 51-59, Apr., 2006.
[18] M. D. Peña-Bandalaria, “Impact of ICTs on open and distance learning in a developing country setting: the
Philippine experience,” International Review of Research in Open and Distance Learning, vol. 8(1), pp. 3145, Feb., 2007.
[19] I. Hussain, “A study of student’s attitude towards virtual education in Pakistan,” Turkish Journal of
Distance Learning, vol. 8(2), pp. 69-79, Jan., 2007.
[20] T. Belawati, A. Zuhairi, “The Practice of a Quality Assurance System in Open and Distance Learning: A
case study at Universitas Terbuka Indonesia (The Indonesia Open University),” International Review of
Research in Open and Distance Learning, vol. 8(1), pp. 46-60, Feb., 2007.
nd
Congruent Networking: A novel concept and model of recommendation system for utilizing
the electronic social networks to maximize quality of service in a context
Nitin V Pujari
Department of Computer Science
PES Institute of Technology Bangalore, Karnataka, India
[email protected]
Vinay S K
[email protected]
Vishal V K
[email protected]
Abstract:
Ever since the dawn of the civilization, every living entity on the planet, spends most of the time to
survive effectively by forming its own network consisting of homogenous and / or heterogeneous entities. This
can be termed as a social network in the context of that entity. Hence social networks together form groups,
communities, societies and civilizations. These networks exhibit the property of homogenous and heterogeneous
interests during their formation. A basic congruence does exist when such social networks are formed. The
utilization of such social networks to carry out specific tasks needs the formation of a sub-network, which hence
can be termed as a congruent network. Therefore a Congruent Network can now be defined as a network of
entities specifically formed to carry out a task effectively. The strength of a congruent network is directly
proportional to the sum of the degree of congruence of each entity selected for that context. This work attempts
to propose a concept and model for generating generic and specific effective congruent networks for a context
and / or a spectrum of contexts using the current electronic social networks. This work also proposes
representation system of the congruent networks, and how congruent networks can be used as recommendation
systems.
Keywords-components;
social
networks;
context;
congruence;
congruent
network;
homogeneous;
heterogeneous
I. INTRODUCTION
It is known that human is a social animal [1], meaning that he/she would be highly interactive with
other members to the point of having a recognizable and distinct society. We survive in the ontology of the
world [2] around us by interacting with other entities of the society. Our interaction with others is not only the
key to our survival, but also to our happiness and the success of our tasks. As explained by Aristotle [3],
nd
“Anyone who either cannot lead the common life or is so self-sufficient as not to need to, and therefore does not
partake of society, is either a beast or a god”. As we form a Society, we are all connected in one way or the other
with whom we like and also whom we hate, just as it is often said that “the friend of my friend is my friend” and
“the enemy of my enemy is my friend,” so too the friend of my enemy is my enemy. This implies that, we have
a social network of our own. We can derive implications of social network as rules [4]. Some of the rules are
Rule 1: We shape our network
Rule 2: Our network shapes us
Rule 3: Our friends affect us Rule 4: Friend of a friend may affect a friend Rule 5: The network has a
life of its own
Now there are service providers who offer services for the people to form or to be a part of a social
network .The features of these services enable one's social network to grow exponentially. Attempts are also
being made to make the social network as hybrid as possible. The social link [5] is being formed on various
parameters such as hobbies, folksonomy, profile, views etc. As per rule 5, and factors for creating social
networks, lifetime of social network varies and it is stronger in particular contexts [6]. Context here means we
and members of our network are connected together effectively in a circumstance to do a task.
Earlier investigations carried out discuss the community of purpose and communities of interest are
formed voluntarily based on an intent or purpose which sometimes may mask the forced intent [7][8]. Whereas
congruent networks once formed will probably be more suitable to carry out the purpose or interest effectively
because most of times it considers attributes that are derived from sub-conscious actions as well. Ego-centric
congruent networks may be used to check the degree of congruence w.r.t community of purpose and community
of interest.
The intent that connects us stronger in a network can be termed as “Congruence”. Such networks can
be termed as “Congruent networks”. The dictionary meaning of congruence is as follows “The quality of
agreeing; being suitable and appropriate”. The mathematical meaning of congruence can be stated as two figures
or objects are congruent if they have the same, shape and size, or if one has the same shape and size as the
mirror image of the other.
II. MOTIVATION
The underlying assumption to have congruent networks is that “humans think alike, act alike in some
situations”[9]. We might have observed that, we and our friends react in a same way in some contexts. If we
make our network with friends who are congruent with us, our social actions can be decided productively. With
the increasing use of internet, most of our communication in a network happens online. Social networking sites
are available to support social interaction and most of their service offerings are free. There has been a dramatic
increase in the number of internet users, most of them are spending time on social networking sites.
nd
Figure 1: Number of users of Facebook
Figure 2: Number of Twitter users
As depicted by Figure 1, Figure 2(Source: Forbes, Twitter, Facebook), a question arises. Are current
human entities becoming more social? This question can be answered in 2 ways. Firstly, people are becoming
more active in social networking sites, thereby making them more social online; secondly, Human is not
becoming socially connected but impulsively connected.
Even though the social networking sites are providing various benefits, they have limitations of their
own. Each of them provide a particular set of services and are used for different purposes. But most of them
provide minimal services that can be considered beneficial for both the stake holders- user and service provider.
Congruent networks thus may provide a better supplemented solution by generating suitable recommended
networks for specific and/or generic contexts. Contexts can be task(s), activity(s), event(s), purpose(s) etc.
III. CONGRUENT NETWORKS BEYOND SOCIAL NETWORKS
A. Relationship between social networks and congruent networks
Congruent networks are subsets of networks derived from social networks as depicted below in Figure 3.
Figure 3: Social Networks and Congruent Networks in social context
Figure 3 depicts that, to form congruent networks it is necessary to have associated social networks.
The creation of congruent networks is influenced by the typical characteristics of human behavior of thinking
similarly with respect to the other, in a context or a situation. It gives us a new dimension to view congruent
networks outside the social context, if we associate biological congruence as well. Congruent networks can also
be viewed as networks that utilize various aspects of human behavior including their biological composition in a
context which otherwise is not within the spectrum of social networks. This assumption now makes congruent
networks a superset of social networks which is the complement of the concept depicted in figure 4.
nd
Figure 4: Social Networks and Congruent Networks in universal context
B. Limitations of Social Networks
Even though social networks are large, growing fast, they have many limitations, and some are listed below.
1) Slowing down development of interests: They are good for maintaining ties established between people
before, but they don't help to create new ties. They close people in particular communication circle and
slowdown the development of their interests.
2) Changing expectations about privacy: Sharing is the most common activity in social networks. Even though
it is not bad to share good things (creating awareness), same thing cannot be applied everywhere. For
example, confidential information should not be shared in any way. Some specific limitations are :
•
Increased usage of Social Network Analysis by companies
•
Long term changes in cultural norms
3) Higher speed of interaction creates impatience and expectations of virality: If one can now share something
with a thousand people in a minute, things move much faster. This can be a good thing, but it also means that
news and information can spread before people have time to carefully process it and reflect upon it, and
things can also fade out more quickly. This spontaneity, good and bad, leads to higher impatience. People
increasingly expect things to "go viral", and if it does not, it is viewed as a failure. If one composes a song
and share it and if it does not become popular, then it may result in a judgement that the song is not good.
This expectation of virality also creates strong incentives within all industries, and specifically within the
technology industry, to rely on virality to promote products. User acquisition is difficult and often expensive,
and hence virality seems to be a boon: The downside of this is “a new unwillingness to create products for
private use”.
4) Decreased ability to recover from mistakes: If one makes a mistake during their childhood, it is remembered
longer, and could have serious long term repercussions. It is also harder to change one s mind and evolve
their own views without being labelled as a "flip flopper". As one gets older, their views naturally become
more informed and more nuanced. Evolution should be considered as a sign of maturity. This can also lead
to more immediate and shallow assessments of people[11].
5) Increased social comparison, potentially leading to envy and loneliness: People naturally compare
themselves to other people. However, instead of just comparing oneself to the people seen in day to day life,
we can compare ourselves to everyone else in our social network[12]. People (tend to) promote a positive
image of themselves on social networks. With exceptions, they avoid posting many messages of loneliness
or isolation, pictures where they look unattractive, etc. Not only is one comparing himself/herself to other
people, he/she is also comparing himself/herself to the positive and flattering portrait of themselves that
nd
people like to project. Economic research has shown that after a certain absolute level of wealth, the relative
wealth impacts our happiness more than the absolute level.
6) Increased prevalence of "groupthink" and prioritization of social approval: The products built around the
social networks are based on the idea that people want the guidance of others in making decisions, or at least,
the approval of others. Many applications or products provide the "value" of knowing popular opinion. It is
more likely that we want to know at which restaurant our friends are eating so that we can also eat along
with them. "Popular" is not the same as "good", which is something that seems to get lost in social
networks. Maybe we don't actually like what everyone else likes. At the most, we miss out on trying
something we would have liked, because none of our friends tried it earlier. At the least, we feel secluded
and isolated because our preferences are completely different from those of other people in our network, and
the only content that gets made is that which appeals to the broadest mass market[10]. This is amplified at
times, when we want the approval of others. We cannot decide on our own even for our personal life. People
always involve society.
More importantly, large innovations come from doing what other people are not doing, from
questioning and pushing on conventional wisdom. Huge victories in social justice comes from people standing
up and saying "I know this is how we have always done things and I know this is what our societal norms are,
but I think they are wrong anyway." People should be willing to do what is not popular.
7) Addiction: Marriages are broken because people pay more attention to social networks or games , than paying
attention to each other. That is a larger cost to humans than anything that has happened due to privacy or
groupthink[10][11][12].
C. Solutions: Congruent Networks
From the discussion above, an attempt to solve all of the above limitations plus other related limitations
is made by creating “Congruent Networks”. There are 2 major things that contribute towards the solution. They
are
1. Context: Our actions are limited by the context. Contexts are created to benefit all the stakeholders involved
bilaterally. Mutual benefit helps in strengthening our ties with our contexts.
2. Only Social: We are considering only social attributes and social views. It can be shared only in the contexts
which are meaningful to us. Privacy is not an issue here.
D. Difference between Social Networks and Congruent Networks
nd
TABLE I. SOCIAL V/S CONGRUENT
IV.
WORKING OF CONGRUENT NETWORKS
As explained earlier, it is a must to have social networks before generating congruent networks. With
these social networks, context matching is made. Only the nodes that match the context are retained and
remaining are discarded. To do this, clustering is done with the attributes of the context. Nodes closer and nearer
to each other in the “context space” are retained. Further, with the remaining nodes, congruence value for each
of the nodes with the given context is calculated. Nodes having congruence factor
factor greater than a threshold are
considered congruent with the context.
A. Proposed Architecture for Generation of Congruent networks
Figure 5: Architecture for Generation of Congruent Networks
The major phases of generating congruent networks are identified as follows:
1. Analyzing and optimizing social networks
2. Defining the context
3. Context matching
4. Calculating the degree of congruence
5. Generating congruent networks
nd
1. Analyzing and optimizing social networks: We have many social networking sites where users share some
information which he/she likes with their individual networks (friends). Users are usually identified with
their profiles. User actions are analyzed and recorded in a format called as “User description format (UDF)”.
UDF, along with the result of analysis, also contains profile data. It is important to consider that the user
knows what information of him/her can be shared and it is completely under his/her discretion. Forming a
congruent network does not necessarily mean “sharing of information”. We consider sharing as an action
where information of the user is made available to other parties on the individual's choice. We are only
considering the attributes for creation of congruent networks by maintaining privacy. The focus of the model
is to generate a network of congruent people, without sharing their data. Only the relevant information, that
assists the creation of congruent network in a context, is analyzed and/or created and then utilized.
The social network data collected may have redundant instances of users, their followers or similar
users. UDF takes care of the redundant data, and represents it as a numerical value after collecting, analyzing the
social network data. For example, let us consider the example of Twitter. Instead of storing all the tweets a user
has made, it is better to keep only the recent tweets and keep the classification/domain count for the rest of the
tweets. Classification/domain count of context is a number which indicates the count of actions related to the
context/domain. This demands that, it is necessary to identify the related contexts/domain for each of the users
and the necessary services have to be provided for enabling the change of contexts/domains. The content of
UDF is dynamic, and hence, it can change from time to time.
2. Defining the Contexts: Contexts are defined by properties. Property is an attribute associated with both the
user and context. Attributes can be quantitative or qualitative. Quantitative attributes are measurable whereas
qualitative attributes are not directly measurable but they can be derived through observation and analysis.
Context is defined as a mixture of these attributes with a minimum threshold value for each of the attributes.
Properties of context and attributes of user may not be same. It is possible to match the property of a context
with the attributes of a user. Property of a context can be defined as a combination of attributes. Consider the
following example. It is required to form a team for organizing a social awareness program. The context can
now be defined with the property “Social participation” and with the attributes as a) Interested in social work
b)participated in group activity and c)organized a community development program. Expressing the context
property in terms of user attributes is done by identifying the actions on social networking sites which are
recognizable, recordable and related to the context. For quantitative properties, they are directly expressed as
“required value”. Context can be generic or specific.
Generic Context: Generic context is a context having various sub contexts under it. These contexts are usually
applicable to all the users. “Social Context” is one of the examples of generic contexts. Generic contexts are
defined with multiple properties each of which covers multiple attributes. Generic context takes each attribute as
equally important. One optimization that can be implemented is to consider only attributes that cover sub
contexts.
Specific Context: Specific context has well defined properties and cannot be decomposed further. Specific
context usually captures single purpose.
3. Context Matching: Once the UDF is defined for all the users and contexts are expressed as attributes, context
matching becomes filtering UDFs for contexts. For contexts with only quantitative attributes, matching can be
nd
as simple as an SQL query. For contexts with mixed attributes, matching considers the result of analysis of
social actions and other related attributes. Only the nodes/users satisfying these constraints are retained for the
further process. Context Matching is a kind of clustering where clustering is made on attributes covering the
properties of contexts. The challenges here are i) multiple attributes ii) the order of attributes. One possible
way to handle the above challenges is to perform clustering on distinguishing important attributes and let the
next phase to filter out further.
4. Calculating the degree of congruence: This is the important stage where congruent networks differ from other
electronic social networks. In the previous stage we filter the nodes which do not satisfy the constraints of
context. Here, we consider the “user as the ultimate decision maker” and follow reverse direction to the
previous stage. No matter how important the context is, it is the user who determines the usefulness of the
context. This stage assists the user to determine that importance. Unless the nodes are at centre of the “context
space” where they are congruent already, congruence with the context depends on users. The degree of
congruence can now be calculated in two ways1. Collect inputs from the user and do binary classification for classifying him/her as congruent or non
congruent.
2. Let the user himself/herself agree or disagree with the context.
When all of the non-congruent nodes are filtered out, it is important to give weights to the remaining
nodes with respect to the context. The congruent weight can be calculated as, per node basis or per attribute
basis. Decision again depends on the context and availability of attributes. For a node, the degree of congruence
in the form of weight is calculated as the cumulative sum of all the attributes. For attribute congruence, the
calculation is done for each of the individual attributes.
5. Generation of congruent networks: All the nodes remaining after Stage 4 are congruent with the context. The
context is added to all the nodes in their “Congruence” field of UDF. Adding the context in the nodes has the
advantage of reuse. When the same context is used multiple times, we can directly fetch the congruent nodes
without repeating the process unless new nodes are added due to the dynamic nature of social networks. The
issue to be tackled with this approach is to identify when the node becomes non-congruent with the context.
It can be solved by monitoring the health of each node at regular intervals and update the „congruent
contexts
accordingly.
V. REPRESENTATION AND VISUALIZATION OF CONGRUENT NETWORKS
We can represent a congruent network as a graph, with nodes representing “people” and edges
representing their “congruent relationships”. It is valid to have edges assigned with weights based on the “level
of congruence” between the nodes, because the level of agreement varies for different contexts. These congruent
networks are generic in a way that they cover maximum contexts as possible. These graphs can be colored
differently to differentiate the contexts.
Relationship between the nodes depends on the context. Initially all the nodes are congruently
disconnected. When we consider a particular context, nodes will tend to become congruent with the context.
This relationship between the nodes and context can be represented as a graph, with context as a central node
and all other nodes having edges carrying congruence weight, to the context node.
nd
Figure 6: Representation of nodes and the context
The nodes in these networks will have a lifetime. The lifetime begins, when nodes start to be congruent
with the context and ends when its degree of congruence in terms of weight becomes zero. Networks should be
formed in such a way that lifetime should be same for all the nodes and should not end until explicitly required
by the context.
When we consider a generic context, it becomes difficult to construct and visualize such a graph. Here,
we consider all possible contexts under the given generic context, and all the nodes and construct such graphs
for each of the individual contexts. Finally all these graphs are merged to show all contexts and nodes that are
congruent. For the nodes which are congruent with multiple contexts, multiple edges emerge from the node to
the context nodes, with different colours for different contexts. Such a graph can be visualized as shown below:
Figure 7: Representation of nodes in the generic context
The methodology described above concentrates on generating congruent networks for the given
context. When it comes to the individual nodes/users, it can now be used as recommendation engine. In the first
stage of social network analysis and optimization, we can identify user specific contexts. Based on the result of
analysis and contexts to which he/she is congruent, recommendation of entities for the task required by entity
can be easily done. Consider two entities which have same degree of congruence in context1. First entity has
high degree of congruence in context2 and second entity in context3. When second entity wants to know
something/get benefit from context2, instead of trying out all the entities which are congruent in context2, it can
now directly try with first entity which has the same degree of congruence in context1 and can come to the
decision on how well the context2 benefits him. For example, consider a scenario of virtual learning systems.
We call the entity which wants to learn as 'learner entity' and entity that wants to assist others in learning as
'mentor entity'. Instead of getting help to learn new things from unknown mentor entities, the proposed system
can now be utilized to identify the mentor which has proficiency in the required domain and congruent with the
nd
learner entity. Because these entities are congruent in the context of social learning, the resultant system can be
used as recommendation engine for carrying out effective teaching learning.
VI. CONCLUSION , OPPORTUNITIES AND APPLICATIONS OF CONGRUENT NETWORKS
The work carried out in this paper conclude that the proposed concept and model cane be used by the
existing or upcoming electronic social networks to enhancing their untilization for maximixing quality of
service in a context. This work also opens up the areas that can be chosen for research to apply proposed
congruent network concept and model may include Maximizing team productivity, Assignment problem
,Formation of context based Knowledge groups ,Human relationship management , Comparative study of
behavioural v/s Biological parameters to identify congruence, Product - Entity based congruence, Event
Managed based Congruence, Pre-Market Analysis, Trend based Market analysis, Matrimonial services,
Recruitment Services, E-teaching and E-learning, Graph theory based approach for relationship analytics etc.
REFERENCES
[1] Nancy J. Nersessian, M. Gorman, R. Tweney, & D. Gooding, A. Kincannon, "Interpreting Scientific and
Engineering Practices: Integrating the cognitive, social, and cultural dimensions, In New Directions in
Scientific and Technical Thinking"
[2] John R. Searle, "Social ontology Some basic principles", University of California, Berkeley, USA
[3] Alexa Koenig, "From Man to Beast: Imprisonment and Social Death", University of California, Berkeley School of Law, University of San Francisco, August 20, 2011
[4] Nicholas A. Christakis , James H. Fowler , "Connected : The Surprising Power of Our Social Networks and
How They Shape Our Lives"
[5]
RossanoSchifanella,
Alain
Barrat;
CiroCattuto,BenjaminMarkines,
FilippoMenczer,
"Folks
in
Folksonomies:Social Link Prediction from Shared Metadata", arXiv:1003.2281v1 [cs.CY] 11 Mar 2010
[6] Brian Skyrms and Robin Pemantle, "A dynamic model of social network formation" , special series of
Inaugural Articles by members of the National Academy of Sciences elected on April 27, 1999.
[7] Linda Carotenuto, William Etienne, Michael Fontaine,Jessica Friedman, Michael Muller, Helene Newberg,
Matthew Simpson, Jason Slusher, and Kenneth Stevenson, ," CommunitySpace: Toward Flexible Support
for Voluntary Knowledge Communities", Lotus Institute and Lotus Research, Lotus Development
Corporation, 55 Cambridge Parkway, Cambridge MA 02142 USA
[8] Constance Elise Porter, "A Typology of Virtual Communities: A Multi-Disciplinary Foundation for Future
Research", Journal of Computer-Mediated Communication, Volume 10, Issue 1, page 00, November 2004.
[9] Joan R. Rentsch, Richard J. Klimoski, "Why do „great minds
think alike?: antecedents of team member
schema agreement", Journal of Organizational Behavior,Special Issue: Shared Cognition,Volume 22, Issue
2, pages 107–120, March 2001
[10] Raymond T. Sparrowe, Robert C. Liden, Sandy J. Wayne, and Maria L. Kraimer "Social networks and the
performance of individuals and groups". Acad manage j april 1, 2001 44:2 316-325;
[11] Catherine Dwyer, Starr Roxanne Hiltz, Katia Passerini,"Trust and privacy concern within social networking
sites: A comparison of Facebook and MySpace", Proceedings of the Thirteenth Americas Conference on
Information Systems, Keystone, Colorado August 09 - 12 2007
[12] Kimberly S. Young and Robert C. Rogers. "CyberPsychology& Behavior". SPRING 1998, 1(1): 25-28.
doi:10.1089/cpb.1998.1.25.
nd

Full Citation

Transcription

Similar documents

Sure-Win Dip

This is only a preview. The actual page count of this

Dumex Code violations (Danone)

This is only a preview. The actual page count of this

Qty Price Total Creating the Fit S$19 Please send this form and

22-Nov-2012 - Singapore Gaelic Lions

Property Life Media Kit 2013

ZOUKOUT 2011

PDF - Aetius Capital

Night Festival 2012 Sunsilk Academy Fantasia Dole Yonanas