Towards a Smart (City) Data Science. A case-based retrospective on policies, processes, and agents
0 likes155 views
This document summarizes a presentation by Dr. Enrico Daga on smart city data science. It discusses several projects and initiatives in Milton Keynes, UK related to building infrastructure for smart city applications and research. This includes a data hub cataloguing city datasets, tools to support data science education and pilots with local businesses. It also covers work on privacy-aware systems, policy propagation in data flows, and using city data to power simulations and games.
Towards a Smart (City) Data Science. A case-based retrospective on policies, processes, and agents
- 1. Towards a Smart (City) Data Science Dr Enrico Daga Research Fellow - Knowledge Media Institute - The Open University Confluence-2020 - Amity University, Noida, Uttar Pradesh, India. January 2020 enrico.daga@open.ac.uk | @enridaga | http://www.enridaga.net/ A case-based retrospective on policies, processes, and agents
- 3. A large collaborative initiative (£16M) to support city innovation in Milton Keynes.
- 4. Technical Infrastructure for Research and Development Catalogue of datasets relevant for the city of Milton Keynes Development environment for Smart City applications Environment for Data Science research and education Innovation platform // SMEs pilots http://datahub.mksmart.org
- 5. Data Science with heterogeneity Data sources Licensing Policing Ownerships Provenance Privacy Sharing
- 6. Top MK is a virtual card playing game where each card represents a ward in Milton Keynes, with characteristics such as area, population, level of qualifications, etc. Two players, one human and the other automatic, try to win the other’s cards by choosing the characteristic that has the best chance to win against the other card.
- 7. The problem of exploitability "Data exploitability" is the assessment of the policies associated with the data resulting from the computation of diverse datasets in complex data flows. • Data come from different owners and have different licenses. • Data are processed into new data before being reused. • What are the policies that apply to the output data? Could we commercialise such a game?
- 8. Policy Propagation How to support data managers on making aware developers about the policies an output might derive from the data sources? Three Components: • Licence Model - W3C ODRL • Process Model - The Datanode Ontology • Rules -Policy Propagation Rules http://purl.org/datanode/ns/
- 9. Policy Propagation propagates(duty:attribution,processed into) … has(dataset1,permission:DerivativeWorks) has(dataset1,duty:attribution) … ODRL - License and Terms of Use Datanode - Process Policy Propaga:on Rules has(output, duty:attribution) has(output, permission:commercialise) … has(X,P) ⋀ propagates(P,R) ⋀ relation(R,X,Y) → has(Y,P) These are the policies derived from the sources! http://purl.org/datanode/ns/
- 10. Holistic approach to metadata management
- 12. CityLABS • A place for SMEs to work with academic and industry leaders to develop concepts into prototypes for new products and services. • Urban Business Lab • Tech Design • Prototype Evaluation • Access to the MK Data Hub. • We supported more than 40 SMEs in 2 years
- 13. Impact case: My New Term Eduvocation primary business relates to job advertising in the education settings The developers of MyNewTerm built the first pilot of a geo-spatial search engine with the MK Data Hub (now online!) We are helping them to reuse Open Data (UK Census statistics) to assign quality of life indicators to places and use them to recommend job adverts with relation to teachers’ life interests enrico.daga@open.ac.uk https://www.mynewterm.com/
- 14. http://datahub.mksmart.org/demos/eduvocation/ Quality of Life: “dynamic” indicators
- 15. • 5G connectivity will enable an increased amount of data to be collected in real-time • City Systems can exchange their data to support services • These certainly include personal and sensitive data (e.g. health records can support emergency response) • Relevant information is hard to identify: large amount of heterogeneous data • Not all information is helpful Privacy-aware City Systems
- 17. Privacy-aware City Systems (work in progress) How to reuse live health records to support early emergency response? How to maximise UTILITY and minimise DISCLOSURE? Approach: 1. Synthetic Dataset of Healthcare Records in FHIR (Synthea) 2. Analysis of National H&S guidelines for evacuation Personal Emergency Evacuation Plan (PEEP) 3. FHIR Schema Annotation (ontology-based approach) 4. Reasoning about time and validity (e.g. conditions that are chronic vs temporary) 5. Common sense reasoning (e.g. relying on ConceptNET, Wikidata, …) 6. Classification of need (Mobility, Respiratory, Mental, …) https://github.com/synthetichealth/synthea
- 18. SciRoc is a EU funded project whose aim is to bring robotic tournaments in the context of smart cities. The first international competition took place in Milton Keynes, on 18-21 September 2019. Challenge comprises 5 episodes, testing Human-Robot Interaction, Navigation, Manipulation, Autonomous Flying, Humanoid Robotics and Interaction with smart city infrastructure. http://sciroc.eu
- 20. The role of the MK Data Hub • To simulate the smart city environment during the competition, for example, by providing the Coffee shop Menu in E03 or by keeping track of the status of the inventory in E07 • To configure the benchmark parameters (e.g. change the order for E07) • To monitor (status & location) and report on the internal status of the robots
- 21. To engage with the public, by displaying the status reports from the robots on the screens distributed around the arenas (and on the @mkdatahub Twitter account!), adding an important element of storytelling and explainability to the sometimes unclear behaviours of the robots The role of the MK Data Hub
- 23. The role of the MK Data Hub More than 200k data points were sent by robots during 3 days of trials. This information can be reused • (a) to enable a deeper analysis of the shortcomings of robots, • (b) to compare the performance of the same robot in different trials, and in different competitions among the years, as well as • (c) for archival purposes. Ultimately, recording robot messages and analysing their behaviour may constitute the first step towards benchmarking elements such as self-awareness and deliberation capabilities.
- 25. Towards a (Smart) City Data Science Policy and Process Reasoning • Not only access control! Understanding the process at the knowledge level is key • Data from distributed sensor networks require this management layer • We need a holistic approach to (meta)data management Personalised, adaptable analytic pipelines • One size does not fit all! • Workflow engines just put the problem on users • We need to work on how to make our methods more flexibles. • Knowledge engineering, e.g. problem solving methods have something to say here.
- 26. Towards a (Smart) City Data Science Privacy • Managing who is affected by the disclosure of information is key to enable deployment of distributed (smart city) systems • Need methods for translating information requirements to processing pipelines • We cannot tackle privacy without understanding utility! Explainable Agents • Being able to monitor agents’ intentions is key to (a) assess their capabilities of interpreting the context, and (b) to enable other robots (and city systems) to dynamically cooperate • We need work on how to represent and reason upon their goals and deliberations
- 27. Thank you Towards a privacy-aware information system for emergency response. (2019) Morales Tirado, Alba, Enrico Daga, and Enrico Motta. In Proceedings of the International Conference on Information Systems for Crisis Response and Management, pp. 1411-1413. Propagation of policies in rich data flows. (2015) E Daga, M d'Aquin, A Gangemi, E Motta. Proceedings of the 8th International Conference on Knowledge Capture, 5 A BASILar approach for building web APIs on top of SPARQL endpoints. E Daga, L Panziera, C Pedrinaci. CEUR Workshop Proceedings 1359, 22-32 An incremental learning method to support the annotation of workflows with data-to-data relations. (2016) E Daga, M d’Aquin, A Gangemi, E Motta. European Knowledge Acquisition Workshop, 129-144 Addressing exploitability of smart city data. (2016) E Daga, M d'Aquin, A Adamou, E Motta. 2016 IEEE International Smart Cities Conference (ISC2), 1-6 Reasoning with data flows and policy propagation rules. (2018) E Daga, A Gangemi, E Motta. Semantic Web 9 (2), 163-183 Robot–City Interaction: Mapping the Research Landscape—A Survey of the Interactions Between Robots and Modern Cities. I Tiddi, E Bastianelli, E Daga, M d’Aquin, E Motta. International Journal of Social Robotics, 1-26 Update of time-invalid information in knowledge bases through mobile agents. I Tiddi, E Daga, E Bastianelli, M d'Aquin enrico.daga@open.ac.uk