Semantic Web Applications

1
Semantic Web Applications

• Prof. Dr. Heiner Stuckenschmidt
• Universität Mannheim
• HWS 2007

2
Outline

• The Semantic Web Challenge
• Infomation Portals MultimediaN
• Semantic Browsing COHSE
• Personalized Search MusiDB
• Collaborative Editing Semantic MediaWiki
• Social Network Analysis Flink

3
Semantic Web Challenge

• Contest for the most advanced and most useful
  application of semantic web technologies
• Organized since 2003
• http//challenge.semanticweb.org
• Goal (official) Demostrate the potential of
  semantic web technologies for solving real world
  problems
• Goal (inofficial) Motivate Master and PhD
  Students to do lots of programming work -)
• Participants about 10 20 each Year

4
Waht is a Semantic Web Application?

• Text from the Semantic Web Challenge home page
• A Semantic Web Application has to meet the
  following minimal requirements
• First, the meaning of data has to play a central
  role.
• Meaning must be represented using formal
  descriptions
• Data must be manipulated/processed in interesting
  ways to derive useful information and
• this semantic information processing has to play
  a central role in achieving things that
  alternative technologies cannot do as well, or at
  all
• Second, the information sources used
• should have diverse ownerships (i.e. there is no
  control of evolution),
• should be heterogeneous (syntactically,
  structurally, and semantically), and
• should contain real world data, i.e. are more
  than toy examples.
• Third, it is required that all applications
  assume an open world, i.e. assume that the
  information is never complete.

5
Example 1 MultimeniaN

• Application domain cultural heritage
• Preservation of cultural heritage for the
  following generations
• Digitalization of collections from important
  libraries and museums
• Provide access to objects across different
  collections and locations
• Problems
• Integration of heterogeneous data sources
• Advanced search for objects of a certain style or
  period

6
Data Sources

• Artchive.com online Art Portal(gt3000 Objekte)
• Rijksmuseum Amsterdam (gt700 Objekte)
• Ethnological Museum Amsterdam (gt 23.000 Objekte)
  
• Tropical Museum Amsterdam (gt78.000 Objekte)
• Royal Library of the Netherlands(gt1600 Objekte)

7
Background Knowledge

• Objects from all collections are linked to
  jointly used thesauri and vocabularies
• Getty Arts and Architecture Thesaurus (AAT)
• Union List of Artists Names (ULAN)
• Thesaurus of Geographic Names (TGN)
• Ethnological Thesaurus (SVNC)
• General Thesaurus of the English Language
  (Wordnet)

8
Interface
9
Architectur
10
Benefits of the Technology

• Unified description of art objects from different
  sources and collections
• Facetted search Role of query terms is
  explicated
• Silber as material vs. Silver as part of the
  title
• Background Information about Objects (style,
  period, related artists)
• Explicit relations between objects

11
Example 2 Semantic Browsing

• Application Searching Information
• Looking for infromation on the web either for
  private or for business purposes
• Problems
• The usual ones
• Idea
• Extending existing browsers with the ability to
  use semantic web contents to support search

12
Semantic Web Browsers

• Automatic recognition of objects, persons,
  concepts etc. mentioned on a web page on the
  basis of an underlying ontology
• Support the Annotation of information contents,
  e.g. by proposing annotations and keywords
• Automatic generation of hyperlinks
• To other objects of the same type
• To pages talking about the same concept
• To background information about objects or
  concepts
• Example Piggy Bank, Magpie, COHSE

13
COHSE
14
Benefits of the technology

• Same as above, but
• Applicable to standard web pages, provided that
• A suitable ontology is available
• The terms occurring on the page can be mapped
  into this ontology
• Limited support for personalization by means of
  using special ontologies
• Enrichment of exsiting web resources by
  supporting semantic annotation of contents

15
Example 3 Personalization

• Application
• Proactively suggest relevant contents to the user
• Special Case suggest products the user might
  want to buy
• Problems
• Lack of information about the users interest
• Interests are often ambiguous (z.B. hard music)
• Classification of objects missing or heterogeneous

16
Example MusiDB

• Search for Music
• Music-Ontology Artists, Songs, Albums
  http//musicbrainz.org/
• Example Which Album do I have to buy to get this
  song?
• MusiDB System http//www.cs.vu.nl/rstegers/iwa/

17
Interface
18
What about unknown music?
Which Style do you prefer ?
19
Recommender Systems
Users
2. Feedback
Music
1. Recommendation
3. Improved Recommendation
...
20
(No Transcript)
21
Benefits of the Technology

• As before
• Unique identifiers for Artists etc.
• Explicit Knowledge about relations
• In addition
• Genre-hierarchy as basis for recommendations
• Classification of artists and songs based on user
  feedback

22
Example Collaborative Knowledge Creation

• Users jointly annotate Web pages and create
  knowledge models (a la Web 2.0)
• Prominent examples Social Tagging
• Del.icio.us
• Flikr
• Real Value is added when users jointly create
  more complex knowledge structures (i.e. relations)

23
Example Semantic Mediawiki

• Follows the Wikipedia paradim
• In addition to text, structured information can
  be added using a simple syntax
• Classification of Objects
• Properties of and relations between objects
• Demonstration Domain
• Persons, Conferences

24
Interface
25
Benefits of the Technology

• Large scale acquisition of structured knowledge
• Automatic Generation of Reports and Factsheets
• Semantic Search (e.g. for certain object types)
• Consistent use of type information across
  different languages
• Enables the use outside the wiki by means of
  exporting the knowledge as RDF file, e.g. as a
  basis for analyzing relations

26
Network Analysis

• Use Information about relations between objects
  for futher analyses
• Citation analysis
• Social network analysis
• Duplicate Detection
• Semantic Technologies as basis for such analysis
• Unambiguous representation of relations
• Standard syntax, tool support available

27
Example Flink

• Aanalysis of relations between people involved in
  semantic web technologies
• Data extracted automatically from the web
• Authors of the International Semantic Web
  Conference
• Their publications
• Relations to other authors identified using
  Google
• Analyses
• Important people in the community
• Relevant topics and their interaction

28
Interface
29
Benefits of the technology

• As before
• Unique identifiers for objects
• Explicit relations between objects
• In addition
• Derivation of global properties from relations

30
Conclusions

• Semantic Web Technologies provide means to
  represent strutured information on the web
• unique identifiers
• classifications
• relations between objects
• This representation is an important basis for
  advanced applications
• Search, Personalization, Information aggregation,
  Analyses,