Part 1 - TU Delft
Transcription
Part 1 - TU Delft
29/01/15 The Science of Social Data Geert-Jan Houben TU Delft Web Information Systems & Delft Data Science WIS - Web Information Systems 1 Intelligent typically information technology, computing science, and a natural focus on software Intelligent Cities the complexity is thought to be in efficiency a prescriptive design approach: closed, fixed, centralized data representing the ‘world’ is made to fit the software WIS - Web Information Systems 2 1 29/01/15 Web & Data the Web brought linking data & connecting it to people for inclusion & adaptation: utility a descriptive approach: open, dynamic, decentralized an unprecedented source of data about the ‘world’ (that people are part of) “big data, too big to handle” WIS - Web Information Systems 3 Data Science technology to handle big data asks for a fundamentally new computing science digital (Web) data and its descriptions of the world bring a new complexity to make sense of the data, data science is all about Semantics, w. Scale, Speed & Sustainability WIS - Web Information Systems Data Gap Computational power Time 4 2 29/01/15 Societal challenge of Social Data unlocking data generated by humans and their take on the world, typically with yet-to-discover semantics enabling the effective use of the data for organizations that aim to include and serve large numbers of human users win-win for customers, patients, travelers, students, etc. WIS - Web Information Systems 5 Scientific challenge of Social Data 1) understanding properties of human-generated data 2) creating technology & systems to make sense of data the Web is the largest humanmade artifact ever made and a synergy of technology & humans – scientists & engineers should study fundamentals as well as technology in deployment: utility WIS - Web Information Systems 6 3 29/01/15 Social Data: From the people For the people WIS - Web Information Systems 7 Science of Social Data: Synergy of Technology & Humans Fundamental & Experimental Utility WIS - Web Information Systems 8 4 29/01/15 1 - It’s data ‘about’ people & world ‘About people & world’ most social web data analytics is about re-purposing data that is out there available or we already have in our systems main research goal is to go from low-level features to meaningful concepts WIS - Web Information Systems 9 Interpretation Emergencies & incidents assessing from social web data what is happening or about to happen for public safety asks for well-designed pipelines with many small decisions for producing actionable knowledge requires extensive real-world experimentation WIS - Web Information Systems 10 5 29/01/15 Multiple perspectives Urban analytics seeing the city through open social urban data for modeling & ‘reality checking’ of city flows data can tell different stories from different perspectives and different semantics WIS - Web Information Systems 11 Adaptation Online education analytics to make online education truly learner-centric and to adapt to the students & their backgrounds massive online education is about massively adapting to the context of use with increasing diversity comes importance of social and cultural features: inclusion – Nr.1 TopSector WIS - Web Information Systems 12 6 29/01/15 ‘Data Analytics machines’ Next 1) by design from using data out of manufacturing to generating data about what people want 2) well understood – models, languages & metrics for data analytics to understand & measure its usage and deploy it ‘as a machine’ WIS - Web Information Systems 13 2 - It is data ‘by’ people ‘By people’ besides re-actively taking what’s there, people could contribute to creating digital data – they know best about their world main research goal is to pro-actively stimulate people to effectively create and complement data WIS - Web Information Systems 14 7 29/01/15 Complementing Social sensing engaging people in collecting data about their living environment complementing physical sensing and ‘official’ data WIS - Web Information Systems 15 Collaborative Workforce engagement in the professional environment where bottom up work floor knowledge is not included gamification techniques enable collaboratively building and sharing knowledge that complements professional resources WIS - Web Information Systems 16 8 29/01/15 ‘Data Creation machines’ Next 1) by design sustainable data creation specific for the context 2) well understood – deploy data creation machines with transparent and well-understood properties WIS - Web Information Systems 17 3 - Making sense ‘with’ people ‘With people’ data often in language that humans can understand, while software machines require training & tuning main research goal is to use human computing for complementing software interpretation & making sense WIS - Web Information Systems 18 9 29/01/15 Crowds & Niches Crowd annotation involving general crowds and qualified niches of domain experts for annotating large collections of (heritage) objects with domain-specific expertise WIS - Web Information Systems 19 Domain Knowledge Collaborative knowledge creation for forums and ‘question answering’ platforms where professionals and learners seek domain knowledge technology for interpreting expertise and experts enables people themselves to construct domain knowledge WIS - Web Information Systems 20 10 29/01/15 ‘Human Interpretation machines’ Next 1) by design for sustainable human and crowd enhanced interpretation Crowd capacity: designing a crowd to produce knowledge with given properties 2) well understood – strategies & metrics to effectively & transparently deploy ‘human & crowd interpretation machines’ WIS - Web Information Systems 21 The science of sense making systems research agenda: towards theory and technology for (software and human-enhanced) machines to create value out of data Scale Semantics how to make sense of social data with its variety, accuracy, and diversity? how to handle large volumes of data and engage large groups of people? Speed how to create and interpret in real-time and in changing contexts? Sustainability how to ensure sustained functioning of people-enhanced systems? WIS - Web Information Systems 22 11 29/01/15 Unlocking Social Data: Software & Human-enhanced Machines with Well-Understood Properties WIS - Web Information Systems 23 Science of social data 1. Social data gives us one of the largest reflections of the world, but/and it is a man-made reflection ‘unique opportunity turning into interesting research problem’ 2. Sense & value come from big data, but even more so from what (software and human-enhanced) machines can make of the data ‘V = M * D’ 3. The power of what machines can do with the data needs to be well-understood and transparent for solid engineering and uptake ‘what machines can do and what they cannot do’ 4. Science and technology follow the principles of the Web ‘fundamental & experimental’ WIS - Web Information Systems 24 12