• Save
A Semantic Multimedia Web (Part 2)
Upcoming SlideShare
Loading in...5
×
 

A Semantic Multimedia Web (Part 2)

on

  • 5,563 views

A Semantic Multimedia Web: Part 2 of a series of lectures given at the Universidad Politécnica de Madrid (UPM), May 2009

A Semantic Multimedia Web: Part 2 of a series of lectures given at the Universidad Politécnica de Madrid (UPM), May 2009

Statistics

Views

Total Views
5,563
Views on SlideShare
5,548
Embed Views
15

Actions

Likes
3
Downloads
8
Comments
1

3 Embeds 15

http://www.slideshare.net 9
http://sassou1982.blogspot.com 5
http://www.linkedin.com 1

Accessibility

Categories

Upload Details

Uploaded via as Microsoft PowerPoint

Usage Rights

© All Rights Reserved

Report content

Flagged as inappropriate Flag as inappropriate
Flag as inappropriate

Select your reason for flagging this presentation as inappropriate.

Cancel
  • Full Name Full Name Comment goes here.
    Are you sure you want to
    Your message goes here
    Processing…
  • Great display. I've taken some of the structure graphics together with adapted to my startup
    Sharika
    http://financeadded.com http://traveltreble.com
    Are you sure you want to
    Your message goes here
    Processing…
Post Comment
Edit your comment

A Semantic Multimedia Web (Part 2) A Semantic Multimedia Web (Part 2) Presentation Transcript

  • A Semantic Multimedia Web: Create, Annotate, Present and Share your Media Raphaël Troncy, < [email_address] > CWI, Interactive Information Access
  • Agenda
    • Semantic Web technologies
      • RDF: a general triple model for modeling information
      • RDFS / OWL: languages for representing ontologies on the web
      • SKOS: simple knowledge organization system
    • Modeling your Ontology
      • Methodologies for building ontologies
      • Tools for publishing vocabularies on the web
    • Linked Data
      • Expose you data following 4 basic principles
      • Re-use a growing amount of publicly open datasets
      • Applications powered by linked data
  • RDF
    • is the first layer of the Semantic Web standards
  • RDF
    • stands for R esource D escription F ramework
  • RDF
    • stands for R esource: pages, images, videos , ... everything that can have a URI D escription: attributes, features, and relations of the resources F ramework: model, languages and syntaxes for these descriptions
  • R D F
    • is a triple model i.e. every piece of knowledge is broken down into
    • ( subject , predicate , object )
  • knowledge
    • take for instance the following piece of
  • image.jpg has for creator Raphael and depicts the elephant Ganesh
  • image.jpg has for creator Raphael image.jpg depicts the elephant Ganesh
  • ( image.jpg , creator , Raphael ) ( image.jpg , depicts , Elephant Ganesh ) ( subject , predicate , object )
  • R D F
    • in the atoms of knowledge are triples of the form ( subject , predicate , object )
  • a triple the RDF atom Predicate Subject Object
    • is also a graph model to link the descriptions of resources
    RDF
    • triples can be seen as arcs of a graph ( vertex , edge , vertex )
    RDF
  • Raphael creator image.jpg depicts Ganesh
    • in resources and properties are identified by URIs
    R D F http://mydomain.org/mypath/myresource
  • R D F
    • in values of properties can also be literals i.e. strings of characters
  • http://www.cwi.nl/~troncy#me http://purl.org/dc/elements/1.1#creator http://flickr.com/photos/rtroncy/2923/ http://xmlns.com/foaf/0.1#depicts Elephant Ganesh
  • nodes RDF allows blank
  • anonymous a resource may be i.e. not identified by a URI and noted _: xyz
  • e.g., there exists an image about Ganesh
  • http://comm.semanticweb.org/core#Image rdf:type _:x http://xmlns.org/foaf/0.1/#depicts &quot;Ganesh&quot;
  • break blank nodes the graph, they cannot be reused
  • name your resources and reuse existing names as much as possible
    • < /> has an XML syntax
    R D F
    • don't run away
    • <rdf:RDF xmlns:rdf=&quot;http://www.w3.org/1999/02/22-rdf-syntax-ns#&quot; xmlns:dc=&quot;http://purl.org/dc/elements/1.1/&quot; xmlns:foaf=&quot;http://xmlns.com/foaf/0.1/&quot; > <rdf:Description rdf:about= &quot;http://flickr.com/photos/rtroncy/2923&quot;>
    • <dc:creator rdf:resource= &quot; http://www.cwi.nl/~troncy#me &quot;/> <foaf:depicts> Ganesh </foaf:depicts> </rdf:Description>
    • </rdf:RDF>
    • it's only for ma chi nes
  • open-world assumption as opposed to the closed world assumption of classical systems
  • in short: the absence of a triple is not significant
  • ( image.jpg , creator , Raphael )
    • doesn't mean image.jpg has one creator
  • ( image.jpg , creator , Raphael )
    • means image.jpg has at least one creator
  • if you have no other triples giving authors it does not mean they are not true .
  • RDF S
    • stands for R D F Schema
  • RDF S
    • provides primitives to write lightweight schemas for R D F triples
  • long way
    • a little drop of semantics goes a
  • RDF S
    • provides primitives to...
    • ... define the vocabulary used in triples
    • ... define elementary inferences
  • RDF S
    • to define classes of resources and organize their hierarchy
  • DOCUMENT REPORT
  • RDF S
    • to define relations between resources and organize their hierarchy
    Tom Doc13 AUTHOR
  • CREATOR AUTHOR
  • RDF S
    • allows for multiple inheritance for classes and properties
  • PERSON MAN MALE
  • RDFS
    • relations have a signature
    DOMAIN RANGE
  • RDFS
    • relations have a signature
    • ... the domain is the type of the resource the relation starts from.
    • ... the range is the type of the resource the relation ends to.
  • RDFS
    • relations with several...
    • ... domains mean all domains apply.
    • ... ranges mean all ranges apply.
    • Ranges and domains are optional
  • but looks like object programming,
  • properties
    • are first class citizens,
    • they are not defined inside classes,
    • they have their own hierarchy.
  • no overwriting
    • in particular a property can not be refined for sub classes of its range or domain.
  • multi-instantiation
    • a resource can have several types,
    • it can be the instance of several classes
    • like multiple lights, facets on a resource.
  • RDF S
    • provides primitives to...
    • ... give labels ...
    • ... give comments ...
    • ... for classes and properties
  • RDF...
    • RDFS using the XML syntax for
  • <rdf:RDF xml:base=&quot;http://inria.fr/2005/humans.rdfs&quot; xmlns:rdf =&quot;http://www.w3.org/1999/02/22-rdf-syntax-ns#&quot; xmlns:rdfs=&quot;http://www.w3.org/2000/01/rdf-schema#&quot; xmlns=&quot;http://www.w3.org/2000/01/rdf-schema#> < Class rdf:ID=&quot; Man &quot;> < subClassOf rdf:resource=&quot;# Person &quot;/> < subClassOf rdf:resource=&quot;# Male &quot;/> <label xml:lang=&quot;en&quot;>man</label> <comment xml:lang=&quot;en&quot;>a male person</comment> </Class>
  • <rdf: Property rdf:ID=&quot; hasMother &quot;> < subPropertyOf rdf:resource=&quot;# hasParent &quot;/> < range rdf:resource=&quot;# Female &quot;/> < domain rdf:resource=&quot;# Human &quot;/> <label xml:lang=&quot;en&quot;>has for mother</label> <comment xml:lang=&quot;en&quot;>a female parent</comment> </rdf:Property> </rdf:RDF>
  • RDF S
    • semantics : standard inference rules to derive additional triples from known statements.
  • example of type propagation
    • IF (c 2 , subClassOf , c 1 ) AND (x, type , c 2 ) THEN (x, type , c 1 )
    IF (Man, subClassOf , Person) AND (Tom, type , Man) THEN (Tom, type , Person)
  • example of property propagation
    • IF (p 2 , subPropertyOf , p 1 ) AND (x, p 2 , y) THEN (x, p 1 , y)
    IF (author, subPropertyOf , creator) AND (Tom, author, Report12) THEN (Tom, creator, Report12)
  • example of subClass transitivity
    • IF (c 2 , subClassOf , c 1 ) AND (c 3 , subClassOf , c 2 ) THEN (c 3 , subClassOf , c 1 )
    IF (Person, subClassOf , Animal) AND (Man, subClassOf , Person) THEN (Man, subClassOf , Animal)
  • example of subProp transitivity
    • IF (p 2 , subPropertyOf , p 1 ) AND (p 3 , subPropertyOf , p 2 ) THEN (p 3 , subPropertyOf , p 1 )
    IF (parent, subPropertyOf , ancestor) AND (father, subPropertyOf , parent) THEN (father, subPropertyOf , ancestor)
  • example of domain inference
    • IF (p 1 , domain , c 1 ) AND (x, p 1 , y) THEN (x, type , c 1 )
    IF (author, domain , Human) AND (Tom, author, Report12) THEN (Tom, type , Human)
  • example of range inference
    • IF (p 1 , range , c 1 ) AND (x, p 1 , y) THEN (y, type , c 1 )
    IF (author, range , Document) AND (Tom, author, Report12) THEN (Report, type , Document)
    • OWL is a new layer on top of RDFS
      • OWL Lite
      • OWL DL
      • OWL Full
    • Actually, OWL is a restriction of RDF Classes  Properties  Individuals
  •  
  •  
  •  
  • Semantic Web Tutorials
    • RDF in a nutshell
      • Tutorial ( credits @ Fabien Gandon )
    • RDFS in a nutshell
      • Tutorial ( credits @ Fabien Gandon )
    • Useful Resources
      • http://www.w3.org/2001/sw/
      • http://www.w3.org/People/Ivan/CorePresentations/
  • Agenda
    • Semantic Web technologies
      • RDF: a general triple model for modeling information
      • RDFS / OWL: languages for representing ontologies on the web
      • SKOS: simple knowledge organization system
    • Modeling your Ontology
      • Methodologies for building ontologies
      • Tools for publishing vocabularies on the web
    • Linked Data
      • Expose you data following 4 basic principles
      • Re-use a growing amount of publicly open datasets
      • Applications powered by linked data
  • Modeling Vocabularies/Ontologies
    • Approaches:
      • Ontology learning
      • Ontology reuse
      • Competency questions
    • Methodologies:
      • Uschold and Grüninger
      • M ETHONTOLOGY : whole process of ontology engineering
      • Differential principles
  • Semantic principles
    • Actor
      • Role of a human being
      • Role in making movies
      • Is a character in the movie
    • Director
      • Role of a human being
      • Role in making movies
      • Is not in the movie, but directs it
    Generic “sèmes” Specific “sèmes”
  • Differential Principles Role Actor Director P3 : difference between the sibling nodes P4 : common property shared by the sibling nodes Plays in the movie Directs the movie P1: property with parent node (being a role of a human being) P2 : difference with parent node
  • Differential Principles
    • Amount to explicitly specify what is understood when using a concept
    • Formulate the semantic commitment that should be respected to use concepts
  • Extensional description of concepts Actor Linguistic semantics: Is a role: plays in a movie Denotation: {{w1: d11, d12..}, {w2: d21, d22, …}…} Director: Linguistic semantics Is a role; Directs a movie Denotation: {{w1: d11, d12..}, {w2: d21, d22, …}…} Role of a human being
    • Extensional red links: lattice structure is possible
    • Intensional black links : lattice structure is forbidden (tree)
    {Woody Allen, Clint Eastwood,…} forbidden
  • Summary
  • DOE
  • OilEd
  • WebODE
  • OntoEdit
  • KAON Workbench
  • NeOn Toolkit
  • SWOOP
  • Protégé 4
  • How to support the collaborative ontology modeling and editing task?
  • Knoodl
  • Collaborative Protégé
  •  
  • Neologism SIOC example
  • Assignment
    • VoCamp like idea: let's model some vocabularies ...
      • Brainstorming: pick a domain
      • Choose (essential) terms related to the domain
      • Sketch the model (blackboard)
      • Find and re-use existing vocabularies
      • Publish your model
  • Agenda
    • Semantic Web technologies
      • RDF: a general triple model for modeling information
      • RDFS / OWL: languages for representing ontologies on the web
      • SKOS: simple knowledge organization system
    • Modeling your Ontology
      • Methodologies for building ontologies
      • Tools for publishing vocabularies on the web
    • Linked Data
      • Expose you data following 4 basic principles
      • Re-use a growing amount of publicly open datasets
      • Applications powered by linked data
  • Wake Up Video
  • Linked Data Principles
    • Tim Berners Lee [2006] ( Design Issues )
      • Use URIs to identify things (anything, not just documents);
      • Use HTTP URIs – globally unique names, distributed ownership – so that people can look up those names;
      • Provide useful information in RDF – when someone looks up a URI;
      • Include RDF links to other URIs – to enable discovery of related information
  • Linked Data Issues
    • Issues:
      • These are principles, not implementation advices
      • Many things (deliberately?) kept blurry
      • Non-information resource vs information resource debate
        • Architecture of the World Wide Semantic Web [ AWWSW ]
        • Ontology of Resources for Linked Data [ Halpin & Presutti ]
      • Content Negotiation and 303, httpRange-14 resolution
    • Formats:
      • HTML + RDF/XML vs. RDFa
  • The limit of content negotiation
    • Is conneg-ing between rdf+xml and video/* a good behavior?
      • Current configuration of jigsaw [ W3C Photo RDF ]
      • Equivalent: « Content is &quot;equivalent&quot; to other content when both fulfill essentially the same function or purpose upon presentation to the user » [ WAG ]
      • Abuse of content negotiation: what is the sameness of two representations of a resource?
    • How to draw the line?
      • Are an image and its ASCII version close enough to be content negotiated?
      • Is a transcoded media content still a fragment?
  • The Web of Data
    • Expose open datasets in RDF
    • Set RDF links among the data items for different datasets
    • Over 4.5 billion triples, 5 millions links (March 2009)
    • ... still counting
  • DBpedia
    • DBpedia is a community effort to:
      • extract structured &quot;infobox&quot; information from Wikipedia
      • interlink DBpedia with other datasets on the Web
  • Scrapping infobox data http://dbpedia.org/resource/Madrid http://dbpedia.org/page/Madrid http://dbpedia.org/data/Madrid.rdf
  • Automatic Links Among Open Datasets Processors can switch automatically from one to the other …
  • Linked Data Cloud – March 2007
  • Linked Data Cloud – August 2007
  • Linked Data Cloud – March 2008
  • Linked Data Cloud – September 2008
  • Linked Data Cloud – March 2009
  • Searching Entities in the Cloud
  • Searching Linked Data
  • Linked Data goes mainstream
  • Cultural Heritage Data Cloud
  • Professional Art Annotation with Thesauri from the Web
  • Take Home Message
    • Reuse what is there
      • Of course, one could create RDF data manually … … but that is unrealistic on a large scale
      • Goal is to generate RDF data automatically when possible and &quot;fill in&quot; by hand only when necessary
        • service to get RDF from flickr images http://www.kanzaki.com/works/2005/imgdsc/flickr2rdf
        • service to get RDF from XMP http://www.ivan-herman.net/cgi-bin/blosxom.cgi/WorkRelated/SemanticWeb/xmpextract.html
    • Expose what you make
  • Thanks for your attention http:// www.cwi.nl/~troncy