Data

• Data is everywhere (personal, government, health, events..)
• Uncovering insights
• Predictions
• Making decisions (e.g., where to save energy)
• Smarter systems

We are dealing with

• Decentralized system (data is spread out)
• Data is heterogeneous in nature
• Writing custom APIs (for each data source) is insane!

RDF Model

• Similar to Entity-Attribute-Value (EAV) data model e.g., Sarven.height = 1.65m
• To describe concepts and their relationships (many Graphs)
• Names for resources (URIs)

Human languages

How do we express ourselves?

• Sentences: subjects and predicates (verbs), and sometimes objects
• Body language meh ;)

Vocabularies

• Definitions for the relationship between things e.g.,Friend of a Friend (FOAF), Dublin Core Terms
• Decentralized invention (anyone can make theirs)

RDF Triples

    <Subject> <Predicate> <Object> .

• Different ways to write it
• Serializations: RDFa, Turtle, N-Triples, RDF/XML, JSON

N-Triples (also valid Turtle) example

• Sarven's height is 1.65m.

  <http://csarven.ca/#i> 
    <http://dbpedia.org/property/height>
      "1.65m" .
  

• Montreal is a city.

  <http://dbpedia.org/resource/Montreal> 
    <http://www.w3.org/1999/02/22-rdf-syntax-ns#type>
      <http://dbpedia.org/ontology/City> .
  

RDFa example

• Sarven is 1.65m.

  <p about="http://csarven.ca/#i">
      Sarven is <span property="dbp:height">1.65m</span></p>
  

• Montreal is a city.

  <p about="dbr:Montreal">
     typeof="dbo:City">Montreal is a city.</p>
  

• Sarven is in Montreal.

  <p about="http://csarven.ca/#i">
      Sarven is in <a rel="foaf:based_near"
                      href="dbr:Montreal">Montreal</a>.</p>
  

RDF/XML example

• Montreal is a city.

  <rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#">
    <rdf:Description rdf:about="http://dbpedia.org/resource/Montreal">
      <rdf:type rdf:resource="http://dbpedia.org/ontology/City">
    </rdf:Description>
  </rdf:RDF>
  

Overview of the RDF formats

• Turtle: compact, human readable, triple pattern used in SPARQL
• N-Triples: verbose, small footprint (inspection, processing), exchange
• RDF/XML: verbose, existing toolchains, tree model
• RDFa: (X)HTML page, for humans and machines, can be treated like an API
• JSON: no standard serialization, client-side dev, tree model

SPARQL query example

• Select a list of subjects that is a city.

      SELECT ?city
      WHERE {
          ?city a <http://dbpedia.org/ontology/City> .
      }
      

SPARQL query result example

SPARQL query example

• People who were born in Montreal before 1900

• PREFIX dbo: <http://dbpedia.org/ontology/>
  
  SELECT ?name ?birth ?death ?person WHERE {
       ?person dbo:birthPlace :Montreal .
       ?person dbo:birthDate ?birth .
       ?person foaf:name ?name .
       ?person dbo:deathDate ?death .
       FILTER (?birth < "1900-01-01"^^xsd:date) .
  }
  ORDER BY ?name
      

See results from DBpedia

SPARQL query: federated

• CONSTRUCT {
  ..
  }
  WHERE {
      { SERVICE <http://data.linkedmdb.org/sparql> {
              SELECT * WHERE { .. }
        }
      }
      UNION
      { SERVICE <http://dbpedia.org/sparql> {
              SELECT * WHERE { .. }
        }
      }
  }
  

Where to start?

• Ask good questions!
• Familiarize yourself with data
• Discover and browse data (e.g., see LOD coud)
• Consume and link to others

Make your data discoverable

• Name your resources and publish them
• VoID (describes linked datasets)
• Google Rich Snippets (uses RDFa)
• D2RQ (Relational DB to RDF mapping)

Tools and services

• SPARQL server and endpoint: Fuseki
• Query SPARQL endpoints: DBpedia, World Bank, .. many more, and their availability.
• Linked Data browsers (e.g., Tabulator)
• Command-line analysis: Raptor utils
• Data transformation: Google Refine + RDF extension
• W3C validator