Semantics Perspective on Physical-Cyber-Social Computing
Download as PPTX, PDF1 like717 views
The document discusses the role of semantics in physical-cyber-social (PCS) computing. It addresses challenges in semantic integration and representation when combining data from physical sensors, social media, and cyber networks. Specifically, it examines how to abstract sensor data and contextualize information at different levels, from low-level sensor readings to higher-level interpretations and representations that incorporate contextual metadata like location, time, and situation. The goal is to better understand and utilize context to improve semantic meaning and applications in PCS computing.
Semantics Perspective on Physical-Cyber-Social Computing
- 1. Institute for Web Science & Technologies – WeST Semantics Perspective on Physical-Cyber-Social Computing Steffen Staab, With research work from WeST Team
- 2. Steffen Staab 2Semantics Perspective Prelude: Web Observatory Collect PCSC data & ground truth! Where is our PCSC observatory? Understanding Collecting Describing Analyzing Modeling Predicting Repeating!
- 3. Steffen Staab 3Semantics Perspective And now for something completely different... Where is our PCSC observatory?
- 4. Institute for Web Science & Technologies – WeST Semantics Perspective on Physical-Cyber-Social Computing Steffen Staab, With research work from WeST Team
- 5. Steffen Staab 5Semantics Perspective Physical-Cyber-Social Computing Physical Sensors Social Computing Cyber Understanding And here some miracle M occurs! Let‘s analyse this in more detail!
- 6. Steffen Staab 6Semantics Perspective Physical-Cyber-Social Computing Physical Sensors Social Computing Cyber Understanding And here some miracle M1 occurs Another miracle M2 occurs
- 7. Steffen Staab 7Semantics Perspective Physical-Cyber-Social Computing Physical Sensors Social Computing Cyber View Semantic integration Semantic interfaces Integration Model Understanding Semantic
- 8. Institute for Web Science & Technologies – WeST Semantic Integration
- 9. Steffen Staab 9Semantics Perspective Semantic Integration Ontology definition Ontology mapping Programming with ontologies! (often forgotten part!) Established, but still not known too well Too much myth?
- 10. Steffen Staab 10Semantics Perspective Semantic Integration: refering to a story by Alan Rector
- 11. Steffen Staab 11Semantics Perspective Semantic Integration: refering to a story by Alan Rector FeverElecHeatElemmeasuringTempPhysSensor Complex concept
- 12. Institute for Web Science & Technologies – WeST Semantic Models Also a way to do integration!!!!
- 13. Steffen Staab 13Semantics Perspective Person sensors Physical Sensors Situational Context Environment s. Tool (car...) Group Social Computing Social Context Role ... Cyber
- 14. Steffen Staab 14Semantics Perspective Person sensors Physical Sensors Situational Context Environment s. Tool (car...) Group Social Computing Social Context Role ... Text Cyber Networks Media Actions/Transactions Semantic Represen. Pragmatic interpretat.
- 15. Steffen Staab 15Semantics Perspective Meaning: Contextualization of Topics - MGTM Flickr photos of food, having tags Geographical Topic Models Spatial context improves content analysis
- 16. Steffen Staab 16Semantics Perspective Neighbour Topic Exchange
- 17. Steffen Staab 17Semantics Perspective Neighbour Topic Exchange
- 18. Steffen Staab 18Semantics Perspective Neighbour Topic Exchange Can be generalized to other features than geolocation: • Time (also cyclic time) • Temperature • ....
- 19. Steffen Staab 19Semantics Perspective Meaning: Understanding Context – EU Live+Gov Urban Maintenance (Buitenbeter) Citizens reporting infrastructure damages Text+Image+Sensor readings submitted by citizen • Activity recognition: walking, cycling,.. • Higher level context: shopping, commuting,...
- 20. Steffen Staab 20Semantics Perspective Meaning: Linking of content - EU EPPICS Use Case: Water management Floods of the river Po (Italy) Observations: Physical: • Water level sensors • Smartphone geo positions (where do people drive?) Social & content • What do people tweet („road blocked“, „went there and there“) Challenges: Linking data Correlating data („level X blocks road Y“)
- 21. Institute for Web Science & Technologies – WeST Semantic Views
- 22. Steffen Staab 22Semantics Perspective Relevance in "typical" social network graph
- 23. Steffen Staab 23Semantics Perspective Relevance: Retrieving Node A's news stream
- 24. Steffen Staab 24Semantics Perspective Relevance: Retrieving Node A's news stream
- 25. Steffen Staab 25Semantics Perspective Meaning: Turning sensing of attention into semantics Tagging by sensing Interestingness Content tags to regions
- 26. Institute for Web Science & Technologies – WeST Conclusion
- 27. Steffen Staab 27Semantics Perspective Challenges in Semantics4PCS-computing Sensor ontology (SSN) Better understanding of what does not work Better understanding of whether the problem is in the ontology or in the modeling! Abstraction and contextualization from lowest level interpretation („34734893“) to mid level interpretation • „deviating from the norm by twice standard deviation“ • „cycling“ „running“ To high level interpretation • „broad flooding“ • „shopping“ „commuting“
- 28. Steffen Staab 28Semantics Perspective Conclusion: Semantics in PCS-Computing Representing contextualized facts Extended provenance -> who, where, when,... Understanding context More than sensor readings -> Reporting Using context Influences semantics Influences use of semantics (pragmatics) -> Actions!
- 29. Steffen Staab 29Semantics Perspective Some of many open questions Which context is most useful when? What are good models for higher level interpretations? Notion of pragmatics? Which data is good for which purpose?
- 30. Institute for Web Science & Technologies – WeST Thank you!
- 31. Steffen Staab 31Semantics Perspective References 1. R. Pickhardt, T. Gottron, A. Scherp, S. Staab, Jonas Kunze. Efficient Graph Models for Retrieving Top-k News Feeds from Ego Networks. In: SocialCom 2012 – Proceedings of the 2012 IEEE Fourth International Conference on Social Computing, Amsterdam, NL, September 3-6, 2012. 2. Tina Walber, Chantal Neuhaus, Ansgar Scherp, Steffen Staab, Ramesh Jain. Creation of Individual Photo Selections: Read Preferences from the Users' Eyes. In: Proceedings of ACM Multimedia 2013, Barcelona, October 2013. 3. C. Kling et al. Detecting Non-Gaussian Geographical Topics in Tagged Photo Collections. Tech Report. 4. http://liveandgov.eu/