Introduction to Linked Data for Alchemists

Talk given at the Elixir-Berlin Meetup

Transcript

1. Introduction to Linked Data for Alchemists Marcel Otto March 14, 2019
2. Agenda 1. Whirlwind tour of Linked Data 2. The good, the bad and the ugly 3. RDF.ex
3. Semantic Web a certain vision for the next web a body of standards and technologies an academic research eld
4. Vision
5. Technology Stack
6. Linked Data
7. Linked Data principles §1 “Use URIs to name (identify) things.”
8. Linked Data principles §2 “Use HTTP URIs so that people can look up those names.”
9. Linked Data principles §3 “When someone looks up a URI, provide useful information, using the standards (RDF, SPARQL).”
10. RDF
11. “Berlin is the capital of Germany.” is the capital of Berlin Germany
12. RDF triples http://example.com/capitalOf http://example.com/Berlin http://example.com/Germany predicate subject object
13. RDF literals untyped literals language-tagged literals typed literals "example" "Beispiel"@de "42"^^xsd:integer
14. XSD datatypes
15. RDF graph example http://example.com/capitalOf http://example.com/name http://example.com/population http://example.com/Berlin http://example.com/Germany "Berlin" "3575000"^^xsd:integer
16. RDF serializations N-Triples Turtle JSON-LD RDF-XML RDFa CSVW …
17. N-Triples <http://example.com/Berlin> <http://example.com/name> "Berlin" . <http://example.com/Berlin> <http://example.com/capitalOf> <http://example.com/Germany> . <http://example.com/Berlin> <http://example.com/population> "3575000"^^<http://www.w3.org/2001/XMLSchema#integer> .
18. Turtle @prefix ex: <http://example.com/> . ex:Berlin ex:name "Berlin" ; ex:capitalOf ex:Germany ; ex:population 3575000 ; ex:district ex:Mitte, ex:Kreuzberg ; # ... .
19. Turtle @prefix : <http://example.com/> . :Berlin :name "Berlin" ; :capitalOf :Germany ; :population 3575000 ; :district :Mitte, :Kreuzberg # , ... .
20. JSON-LD JSON example: { "id": 1, "name": "Berlin", "capitalOf": {"id": 2}, "population": 3575000 }
21. JSON-LD { "@id": "http://example.com/Berlin", "id": 1, "name": "Berlin", "capital_of": {"id": 2}, "population": 3575000 }
22. JSON-LD { "@id": "http://example.com/Berlin", "@context": { "name": "http://example.com/name", "capital_of": "http://example.com/capitalOf", "population": "http://example.com/population" }, "id": 1, "name": "Berlin", "capital_of": {"id": 2}, "population": 3575000 }
23. JSON-LD { "@id": "http://example.com/Berlin", "@context": { "name": "http://example.com/name", "capital_of": "http://example.com/capitalOf", "population": "http://example.com/population" }, "id": 1, "name": "Berlin", "capital_of": { "@id": "http://example.com/Germany", "id": 2 }, "population": 3575000 }
24. JSON-LD { "@id": "http://example.com/Berlin", "@context": "http://example.com/context.jsonld", "id": 1, "name": "Berlin", "capital_of": { "@id": "http://example.com/Germany", "id": 2 }, "population": 3575000 }
25. Linked Data principles Rule 4 “Include links to other URIs, so that they can discover more things.”
26. Linking Data { "name": "Elixir Berlin", "link": "https://www.meetup.com/Elixir­Berlin/", "city": "Berlin", "organizer": { "id": 29440972, "name": "Hugo Duksis", } }
27. Linking Data ex:organizer ex:name ex:city ex:Elixir-Berlin-Meetup ex:name "Hugo Duksis" "Berlin" ex:Hugo-Duksis "Elixir Berlin"
28. Linking Data wdt:P1376 ... ex:organizer ex:name ex:city ex:Elixir-Berlin-Meetup ex:name wdt:P1448 "Germany" wdt:P1448 ... wd:Q183 "Hugo Duksis" ... "Berlin" ... wd:Q64 ex:Hugo-Duksis "Elixir Berlin"
29. Linked Open Data Cloud 2007
30. Linked Open Data Cloud 2011
31. Linked Open Data Cloud 2019
32. Consuming Linked Data
33. “Follow Your Nose”
34. SPARQL PREFIX ex: <http://example.com/> SELECT ?capital ?country WHERE { ?x ex:cityname ?capital ; ex:isCapitalOf ?y . ?y ex:countryname ?country ; ex:isInContinent ex:Africa . }
35. More SPARQL highlights CONSTRUCT queries Federated queries SPARQL protocol Reasoning
36. Is the Semantic Web dead?
37. Adoption problems High learning curve Tooling
38. Success stories Bio-medicine Libraries Publishing BBC Nasa IBM Amazon Neptune
39. The case for Semantic Web Open Data Complex data integration problems
40. Knowledge Graphs
41. SEO improvements
42. Solid “Solid (derived from”social linked data“) is a proposed set of conventions and tools for building decentralized social applications based on Linked Data principles.” “The project aims to radically change the way Web applications work today, resulting in true data ownership as well as improved privacy.”
43. RDF & SPARQL in Elixir
44. RDF.ex fully compliant with the RDF 1.1 speci cation
45. URIs RDF.iri("http://www.example.com/foo") import RDF.Sigils ~I<http://www.example.com/foo>
46. URI namespaces defmodule MyApp.NS do use RDF.Vocabulary.Namespace defvocab EX, base_iri: "http://www.example.com/ns/", terms: ~w[Foo bar] end alias MyApp.NS.EX EX.Foo RDFS.subClassOf
47. RDF literals RDF.literal("foo") RDF.literal("bar", language: "de") import RDF.Sigils ~L"foo" ~L"foo"de
48. Typed RDF literals RDF.literal("42", datatype: XSD.integer) RDF.literal(42) RDF.literal(~N[2000­01­01 23:00:07]) RDF.double("42")
49. RDF triples {~I<http://example.com/S>, ~I<http://example.com/p>, RDF.integer(1)} RDF.triple(EX.S, EX.p, 1)
50. RDF descriptions RDF.description(EX.Berlin, [ {EX.name, "Berlin"}, {EX.capitalOf, EX.Germany}, {EX.population, 3575000}, {EX.district, [EX.Mitte, EX.Kreuzberg]} ] ) EX.Berlin |> EX.name("Berlin") |> EX.capitalOf(EX.Germany) |> EX.population(3575000) |> EX.district(EX.Mitte, EX.Kreuzberg)
51. RDF graphs graph = RDF.graph [ {EX.ElixirBerlinMeetup, EX.name, "Elixir Berlin"}, {EX.ElixirBerlinMeetup, EX.organizer, EX.HugoDuksis}, {EX.HugoDuksis, EX.name, "Hugo Duksis"}, ] graph |> RDF.Graph.add({EX.ElixirBerlinMeetup, EX.city, EX.Berlin}) |> RDF.Graph.add(description_of_berlin) |> RDF.Graph.add(another_graph)
52. RDF serializations RDF.NTriples.write_string(graph) # .ttl is the file extension of Turtle files RDF.read_file!("example.ttl") |> RDF.Graph.add(graph) |> RDF.write_file("example.json", format: json_ld)
53. SPARQL.ex graph |> SPARQL.execute_query(""" SELECT ?organizer WHERE { ?meetup ex:name "Elixir Berlin" ; ex:organizer ?organizer . } """) |> SPARQL.Query.Result.get(:organizer)
54. SPARQL.Client # Places with free wi­fi from Wikidata """ SELECT ?item ?itemLabel (SAMPLE(?coord) AS ?coord) WHERE { ?item wdt:P2848 wd:Q1543615 ; # wi­fi gratis wdt:P625 ?coord . SERVICE wikibase:label { bd:serviceParam wikibase:language "[A } GROUP BY ?item ?itemLabel LIMIT 100 """ |> SPARQL.Client.query("https://query.wikidata.org/bigdata/namesp # Result:
55. Resources Introductions (12min) (20min) (3h 49min; paid content, but IMO the best available introduction) Manu Sporny: “What is Linked Data” Markus Lanthaler: “Pragmatic Semantics at Web Scale”, Connected Data London 2018 Brian Sletten: “Resource-Oriented Architectures - Linking Data”, O’Reilly Media, 2015
56. Resources Interesting stuff (12min) (31min) (42min) Ruben Verborgh: “Querying on a decentralized Web”, GraphQL Day 2018 Katariina Kari: “Building, and communicating, a knowledge graph in Zalando”, Connected Data London 2018 Ruben Verborgh: “Solid - taking back the Web through decentralization”, FOSDEM 2019
57. Thanks! Questions?
58. Interested? training for your team support for your projects Contact: http://www.cokron.com info@cokron.com