Linked Data in Use - Front in Bahia 2014

Transcript

1. LINKED DATA IN USE
2. Ícaro Medeiros icaro.medeiros@gmail.com Time de Semântica semantica@corp.globo.com globo.com Front in Bahia 13/9/2014 LINKED DATA IN USE Schema.org, JSON-LD, and hypermedia APIs
3. WHAT IS LINKED DATA?
4. SEMANTIC WEB DONE RIGHT
5. "I have a dream for the web in which computers become capable of analyzing all the data on the web - the content, links, and transactions between people and computers." Tim Berners-Lee
6. ROBERTO CARLOS GOAL AT MARACANÃ
7. http://www.bbc.co.uk/blogs/legacy/radiolabs/s5/linked-data/s5.html
8. http://www.bbc.co.uk/blogs/legacy/radiolabs/s5/linked-data/s5.html
9. Standard Vocabularies • Dublin Core: publications • FOAF: people • SIOC: online communities • DBPedia: Wikipedia in triples
10. Linked Data Principles 1. Use URIs as names for things 2. Use HTTP URIs so that people can look up those names. 3. When someone looks up a URI, provide useful information, using the standards (RDF*, SPARQL) 4. Include links to other URIs, so that they can discover more things.
11. Or… 1. Entity identity 2. Access 3. Structure 4. Integration
12. RDF AND SPARQL
13. Triple representation subject predicate object Neymar is a player Barcelona is a team Neymar plays for Barcelona
14. <Neymar> <Barcelona> <Santos> <Player> <Time> <BarcelonaXSantos:02082013> <Partida> is a is a is a plays for is a home team away team <Camp Nou> location "2013/08/02" date
15. RDF triples @prefix rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>. @prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#>. @prefix foaf: <http://xmlns.com/foaf/0.1/>. @prefix dbpedia: <http://dbpedia.org/ontology> esportes:Neymar rdf:type dbpedia:SoccerPlayer ; rdfs:label "Neymar" ; foaf:name "Neymar da Silva Santos Júnior" ; dbpedia:birthDate "05/02/1992" ; owl:sameAs dbpedia:Neymar . Linking datasets
16. LINKED DATASETS
17. SPARQL query Select all instances of dbpedia:SoccerPlayer class and its names SELECT ?player ?name FROM <http://dbpedia.org> WHERE { ?player rdf:type dbpedia:SoccerPlayer ; rdfs:label ?name . }
18. • Complexity (ontologies?) • NO clear advantage • Triplestore databases: immature • Missing a killer application • Fax problem • Slow adoption Problems
19. LINKED DATA WITHOUT RDF?
20. DON’T YOU WORRY ABOUT A schema.org/Thing
21. SCHEMA.ORG
23. RICH SNIPPETS GOOGLE NOW KNOWLEDGE GRAPH CLEAR ADVANTAGE
24. RICH SNIPPETS City Date Location
26. GMAIL https://developers.google.com/gmail/actions/reference/ fl ight-reservation?hl=pt-BR
27. GOOGLE NOW
28. MAPS
29. SERP
30. <html itemscope itemtype="http://schema.org/Event"> <head> <title itemprop="name"> Front in Maceió </title> crawler saves annotation in KG search results enhanced
31. ONE VOCABULARY TO RULE THEM ALL
32. HOW TO DO IT?
33. <html itemscope itemtype="http://schema.org/Event"> <head> <title itemprop="name"> Front in Bahia </title> ... MICRODATA
34. <html itemscope itemtype="http://schema.org/Event"> ... <section itemprop="location" itemscope itemtype="http://schema.org/Place"> <span itemprop="name">CESMAC</span> <span itemprop="address" itemscope itemtype="http://schema.org/PostalAddress"> <span itemprop="addressLocality"> Maceió </span> ... NESTED PROPERTIES
35. Schema.org usage • Sample 12B pages (Google Index) • 6% domains • ~21% pages • Avg entities/page: 6 • Avg facts/page: 26 http://www.slideshare.net/rvguha/sem-tech2014c
36. JSON-LD
37. Plain old JSON { "name": "Keith Urban", "url": "http://www.keithurban.net/tour/420", }
38. { "@context": "http://schema.org", "@type": "MusicEvent", "name": "Keith Urban", "url": "http://www.keithurban.net/tour/420", } JSON-LD
39. Advantages • Semantic layer on top of JSON • Self-described • Can be added to existing APIs • JSON is extensively used • Data can be anchored in schema.org • Easy serialization to RDF
40. JSON-LD AND HYPERMEDIA APIS
41. GET /events/1 HTTP/1.1 Host: example.com ==================================== HTTP/1.0 200 OK ... Content-Type: application/json Link: <example.com/events/1.jsonld>; rel=http://www.w3.org/ns/json-ld#context; type=application/ld+json { "name": "Keith Urban", "url": "http://www.keithurban.net/tour/420", } Client follows link (hypermedia API)
42. { "@context": "http://schema.org", "@type": "MusicEvent", "@id": "/events/1", "name": "Keith Urban", "url": "http://www.keithurban.net/tour/420", } Self-described Resource
43. { "@context": "http://schema.org", "@type": "MusicEvent", "@id": "/events/1", "name": "Keith Urban", "url": "http://www.keithurban.net/tour/420", "offers": [{ "@id": "/events/1/offer/1", "@type": "Offer" }]} Client follows link Resource+links
44. JSON-SCHEMA
45. GET /events/1 HTTP/1.1 Host: example2.com ==================================== HTTP/1.0 200 OK … Content-Type: application/json; profile=/events/1/_schema { "name": "Keith Urban", "url": "http://www.keithurban.net/tour/420", } Client follows link (hypermedia API) Schema in different resource
46. GET /events/1/_schema HTTP/1.1 Host: example2.com ==================================== HTTP/1.0 200 OK … { "title": "Event", "type": "object", "properties": { "name": {"type": "string"}, "url": {"type": "url"} }, "links": { "rel": "offers", "href": "/events/1/offers" } } Client does not know URLs, just 'rel's
47. HYPERMEDIA APIS
48. Hypermedia? • Don’t remind URLS. Navigate! • Decoupling server and client • Content negotiation • Application treated as State Machine
49. / GET { "items": [ { "title": "News", "@id": "http://example.com/news/", "resource_id": "news" }, { "title": "Sports", "@id": "http://example.com/sports/", "resource_id": "sports" } ], "links": {} }
50. / GET # continuation "links": [ { "href": “http://example.com/“, "rel": "self" }, { "href": "http://example.com/{resource_id}", "rel": "item" }, { "href": "http://example.com", "method": "POST", "rel": "create" }, { "href": "http://example.com/{resource_id}", "method": "DELETE", "rel": "delete" }, { "href": "http://example.com/?page=2", "method": "GET", "rel": "next" } ] …
51. / GET /sports rel:item resource_id:sports GET
52. self /sports/Team/ Barcelona /sports/ inCollection item create delete replace Relations
53. JSON-LD IN HTML
54. JSON-LD annotation <script type="application/ld+json"> [{ "@context": "http://schema.org", "@type": "MusicEvent", "name": "Keith Urban", "url": "http://www.keithurban.net/tour/420", "startDate": "2014-09-07", "performer": [{ "@type": "MusicGroup", "name": "Keith Urban", "sameAs": "http://www.keithurban.net"}], www.keithurban.net/tour Linking datasets Same vocabulary
55. "location": { "@type": "Place", "name": "Sleep Train Amphitheatre", "address": { "@type": "PostalAddress", "addressLocality": "Sacramento", "addressState": "CA", "addressCountry": "US" }}, "offers": [{ "@type: "Offer", "name": "TICKETS", "url": "http://www.keithurban.net/tour/420" }]} //... ] </script> Link is followed by crawler
56. { "@context": "http://schema.org", "@type": "MusicEvent", // ... "offers": [{ "@type": "Offer", "name": "General Tickets", "price": "40.00", "priceCurrency": "USD", "availability": "InStock", "validFrom": "2014-02-15T10:00", "category": "primary", "url": "www.axs.com/events/248328/keith-urban-tickets" }]} keithurban.net/tour/420 ……..
57. www.axs.com/events/248328/keith-urban-tickets
58. JSON-LD IN EMAIL
59. <script type="application/ld+json"> { "@context": "http://schema.org", "@type": "FlightReservation", "reservationNumber": "RXJ34P", "reservationStatus": "http://schema.org/Confirmed", "underName": { "@type": "Person", "name": "Eva Green" }, "reservationFor": { "@type": "Flight", "flightNumber": "110", "airline": { "@type": "Airline", "name": "United" }, //... </script>
60. GMAIL https://developers.google.com/gmail/actions/reference/ fl ight-reservation?hl=pt-BR
61. OPEN GRAPH
64. TOOLS
65. SCHEMA.ORG
66. JSON-LD.ORG
67. g.co/StructuredData
68. http://www.google.com/webmasters/tools/richsnippets
69. STRUCTURED DATA MARKUP HELPER https://www.google.com/webmasters/markup-helper
70. https://www.google.com/webmasters/markup-tester/
72. Semantic annotation goes a looong way • +SEO • +Mobile Rich Info • Applications we don’t even know yet
73. github.com/ globocom/ IWantToWorkAtGloboCom
74. icaromedeiros.com.br • Pelican (Python) • Use ReStructuredText (rst) or Markdown • Generate static pages • Deploy to github pages easily • github.com/icaromedeiros/pelican-microdata
75. .. itemscope:: BusinessEvent :tag: p Today I have presented a talk in :itemprop:`Salvador <location>` at :itemprop:`Front in Bahia <name>`.
76. <p itemscope itemtype="http://schema.org/BusinessEvent"> Today I have presented a talk in <span itemprop="location">Salvador</span> at <span itemprop="name">Front in Bahia</span>. </p>
77. Ícaro Medeiros icaro.medeiros@gmail.com Semantic team semantica@corp.globo.com globo.com slides icaromedeiros.com.br slideshare.net/icaromedeiros