The Web of Data - Tom Heath
- 1. shared innovation The Web of Data Tom Heath Senior Research Scientist Talis Education Ltd July 2012, SSSW2012, Cercedilla, Spain
- 2. shared innovation This talk is about the basics!
- 3. shared innovation Overview • Why a Web of Data is not just about data on the Web • Building blocks of a Web of Data (i.e. Linked Data fundamentals)
- 4. shared innovation Data on the Web
- 5. shared innovation Data on the Web • Takes many different forms • Falls on a continuum of accessibility/usability, structure, meaning and connectedness...
- 6. shared innovation Data on the Web PDF → HTML → Microformats → CSV/XML → Microdata →RDF
- 13. shared innovation Schema.org • The Semantic Web has gone mainstream! • But they're trying to create one ontology for everything! • Is it linked to anything else?
- 14. shared innovation Issues, PDF through Microdata • Structure, parseability – How do we validate what we're extracting? – Is there an underlying data model? • How do we know what the data means? • How do you relate parts of the data together? • How do consumers discover related data?
- 15. shared innovation Data on the Web != A Web of Data
- 16. shared innovation If there are no links, it's not a Web!
- 17. shared innovation The Web of Data
- 18. shared innovation What is the Web of Data? • A boundless graph of interconnected data available online...
- 19. shared innovation What is a Graph? Talis Education Tom employedBy locatedIn Italy studentOf bornIn England tutorAt Enrico 2012-07-08 startedOn SSSW2012 capitalOf locatedIn Spain near Cercedilla Madrid
- 20. shared innovation One Graph, Many Documents Talis Education Tom employedBy locatedIn Italy studentOf bornIn England tutorAt Enrico 2012-07-08 won startedOn SSSW2012 lost Euro2012Final capitalOf locatedIn Spain near Cercedilla Madrid
- 21. shared innovation The RDF Data Model: Triples Object Subject Predicate
- 22. shared innovation The RDF Data Model: Triples Talis Tom employedBy
- 23. shared innovation Why I RDF • Clear underlying data model • Linkability baked into the data model • Ways to express what the data means • Many different ways to write it (for different use cases) • Very easily parsed (in some cases)
- 24. shared innovation RDF Serialisation Formats • RDFa • RDF/XML • N-Triples • Turtle • RDF/JSON
- 25. shared innovation Identifying Things in the Graph • Use URIs • But not any old URIs • HTTP URIs – (almost completely) distributed ownership – anyone (with a credit card) can create/own them – can identify anything – lookup mechanism is baked in
- 26. shared innovation HTTP URIs for Things http://www.talis.com/id/company http://tomheath.com/id/me http://example.org/employedBy
- 27. shared innovation Retrieving Humans over the Web
- 28. shared innovation Linked Data Hands-On Session • Goal: – Give practical experience of working with Linked Data from the Web – Give exposure to different application architectures • Scenario – Building a travel guide application • Tools – Curl, LDSpider, Fuseki, Rapper, Squin
- 29. shared innovation Questions? • Contact Info – tom.heath@talis.com – http://tomheath.com/id/me – http://www.talis.com/ – @tommyh (twitter) • Slides – http://tomheath.com/slides/2012-07-cercedilla- web-of-data.pdf