Loading presentation...

Present Remotely

Send the link below via email or IM

Copy

Present to your audience

Start remote presentation

  • Invited audience members will follow you as you navigate and present
  • People invited to a presentation do not need a Prezi account
  • This link expires 10 minutes after you close the presentation
  • A maximum of 30 users can follow your presentation
  • Learn more about this feature in our knowledge base article

Do you really want to delete this prezi?

Neither you, nor the coeditors you shared it with will be able to recover it again.

DeleteCancel

Make your likes visible on Facebook?

Connect your Facebook account to Prezi and let your likes appear on your timeline.
You can change this under Settings & Account at any time.

No, thanks

Introduction to RDF, Inferencing and Reasoning

Overview talk for Semantic Web languages and reasoning for ESWC Summer School 2011.
by

Spyros Kotoulas

on 22 May 2011

Comments (0)

Please log in to add your comment.

Report abuse

Transcript of Introduction to RDF, Inferencing and Reasoning

LAnguages and Reasoning Spyros Kotoulas
Vrije Universiteit Amsterdam WHAT? WHY? HOW? Implicit information Raki isA GrapeDistilledDrink
GrapeDistilledDrink isA AlcoholicDrink
isA is Transitive

Raki isA AlcoholicDrink Jane is a human
Humans are either men or women
Jane is not a man

Jane is a woman Add value to data Data Integration Combine information
to get better answers Combine information
to detect inconsistencies Backward Forward Simple logics Complex logics Centralized Distributed RDFS subClassof range type inheritance Jane isa woman
woman Subclassof human

jane isa human lists Jane marriedTo john
marriedto domain human

Jane isa human OWL-Horst sameas transitive properties inverse symmetric Jane sameas j. doe
jane marriedto john

j. doe marriedto john OWL full EL- reduce computational complexity QL- enable interoperability with relational databases RL-can be expressed with rules OWL 2 profiles Maximum expressivity Compute closure
i.e. make all implicit
information explicit Infer only information
relevant for a
specific query Fast querying
Simple querying Less index space
Faster loading
Cheap updates Centralised Distributed,
single organisation Fully distributed Single machine
Simpler management
Faster algorithms
Disproportional cost Cluster
Single organisation
Cheaper
Complex implementation Multiple organisations
Multiple data sources
Powerful
Deal with failures
Malicious nodes
Spam Closed Open Low-quality ontological information is bad At best: make your KB inconsistent
At worse: make your KB useless

Need quality control Using well-defined logics
Using formalized knowledge ?a subClassof ?b
?b subclassof ?c

?a subclassof ?c OWL DL Good for expressive ontologies Use description logics Perform inference as part of solving
logic formulas domain Inferencing Reasoning Languages Querying Semantic Web Search engines are great
Extract the most important info
Index and rank billions of Web pages
Figure out what we mean But... it did not answer the query
It shows some web pages where the answer may lie
Relies on user to interpret the content of the pages e.g. the pro athlete who lives in the city with the highest altitude Combine information from several sources Triples consist of:
Subject (URI or Blank node)
Predicate (URI)
Object (URI, Literal or Blank node) URI: <http://τομικρομουπονυ.com>
blahblah:something
Blank node: _:node1
Literal: "Dutch", "Nederlands"@nl,
"4"^^xsd:integer Representation NTriples <http://www.imdb.com/rdf#jjbl> <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> <http://www.imdb.com/rdf#Director> .
<http://www.imdb.com/rdf#jjbl> <http://www.w3.org/1999/02/22-rdf-syntax-ns#label> "Jean-Jacques Beineix" .
<http://www.imdb.com/rdf#jjbl> <http://www.imdb.com/rdf#directorof> <http://www.imdb.com/rdf#bbe> . Turtle @prefix imdb: <http://www.imdb.com/rdf#> .
@prefix rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>
imdb:jjbl rdf:type imdb:Director ;
rdf:label "Jean-Jacques Beineix" ;
"J.J. Beineix" .
imdb:jjbl imdb:directorof imdb:bbe RDF/XML <?xml version="1.0"?>
<rdf:RDF
xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
xmlns:imdb=http://www.imdb.com/rdf#>
<rdf:Description rdf:about=imdb:jjbl>
<rdf:type rdf:resource=http://www.imdb.com/rdf#Director>
<imdb:directorof rdf:resource=http://www.imdb.com/rdf#bbe>
...
</rdf:Description>
</rdf:RDF> Ontology / Schema & SPARQL SELECT { WHERE { ORDER BY ASC/DESC ( ) LIMIT PREFIX basics ?n <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> foaf:Person . ?n 10 foaf: <http://xmlns.com/foaf/0.1/>
myontology: <http://www.mine.com> ?n Used for compactness
as in RDF/Turtle
define as many as you want
name not important Set of variables
must appear on WHERE clause
to select all variables, use *
DISTINCT eliminates duplicates BGPs ?x rdf:type foaf:Person .
?x foaf:name "Manolis" .
?x foaf:knows ?y .
?y foaf:name ?othername . } Optionals ?x rdf:type foaf:Person .
?x foaf:name "Manolis" .
OPTIONAL { ?x foaf:knows ?y .
?y foaf:name ?othername . } CONSTRUCT { ?n myontology:is foaf:Person. Set of variables
must appear on WHERE clause
to select all variables, use *
DISTINCT eliminates duplicates } OR } Unions ?x rdf:type foaf:Person .
{?x foaf:name "Manolis".} UNION {?x foaf:name "M.".} Filters ?x rdf:type foaf:Person .
?x foaf:name "Manolis" .
{?x foaf:knows ?z. }
UNION
{ ?x foaf:knows ?y . ?y foaf:knows ?z . }
?z foaf:first_name "Manolis".
FILTER {?x = ?z} RDF Respositories Where RDF lives
SPARQL is more or less standard
REST API
Do inference Size considerations:
laptop ~10^8-10^9
server ~10^9-10^10
cluster ~10^10 RDF The Web is made for humans Why? A Web for Computers Instance / Data Formats Semantics Access Methods Human I/O Glue Common representation XML RDF RIF REST SPARQL (Semantic Web) Services URI Resolution User interfaces Natural Language Processing Mash-ups XSLTs Extensibility Reuse Ontology of everything Reasoning Ontology/vocabulary alignment Mapping Data mapping Visualisation
Full transcript