The Semantic Web: A network of understanding

1
The Semantic WebA network of understanding

• Jim Hendler
• Univ of MD/RPI
• http//www.cs.umd.edu/hendler

2
Outline

• The Semantic Web
• Past
• Present
• Future

May, 2001
March, 2000
May, 1994
3
Semantic Web hypothesis Heterogeneous Web-based
Information Resources can be connected by
Web-based knowledge models
PVT
Burkitts Lymphoma
Rearrangement of a DNA sequence homologous to a
ltcell-typegtcell-virus junction fragment
lt/cell-typegtin severalltdiseasegtMoloney murine
leukemialt/diseasegt virus-induced
ltorganismgtratlt/organismgt thymomas
PubMed
Oncogene(MYC) Found_In_Organism(Human).
Gene_Has_Function(Transcriptional_Regulation).
Gene_Has_Function(Gene_Transcription).
In_Chromosomal_Location(8q24).
Gene_Associated_With_Disease(Burkitts_Lymphoma).
Semantic Web
PVT
Burkitts Lymphoma
Rearrangement of a DNA sequence homologous to a
cell-virus junction fragment in several Moloney
murine leukemia virus-induced rat thymomas
8q24 PVT1
PubMed
4
Web ontologies

• Web Ontologies are models allowing the linking of
  
• multimedia
• databases
• services
• Web services
• Grid computing
• meta-data repos
• Or any other Web resource!
• Other ontologies
• Anything with a URI

5
The "layercake"
T. Berners-Lee, 2001
6
2001

• Research, experimentation, early demonstrations
• Reminiscent of the early days of the Web

7
2003

• Early government adoption
• Emerging corporate interest

8
2005

• Commercial tools
• Lots of open source software
• Scalability

9
Web Modeling Languages - 2005
10
2006 You Are Here!
11
Significant Corporate Activity

• Semantic (Web) technology companies starting
  growing
• Siderean, SandPiper, SiberLogic, Ontology Works,
  Intellidimension, Intellisophic, TopQuadrant,
  Data Grid,
• Bigger players buying in
• Adobe, Cisco, HP, IBM, Nokia, Oracle, Sun,
  Vodaphone announcements/use in 2005-2006
• Gartner identifies Corporate Semantic Web as one
  of three "High impact" Web technologies
• tools being announced AllegroGraph, Altova,
  TopBraid,
• Government projects in and across agencies
• US, UK, EU, Japan, Korea,
• Life sciences/pharma an increasingly important
  market
• Health Care and Life Sciences Interest Group at
  W3C
• Many open source tools available
• Kowari, RDFLib, Jena, Sesame, Protégé, SWOOP,
  Onto(xxx), Wilbur,

12
Richer metadata
Embedded meta-data
Data harvesting visualization
Enterprise data integration
"Corporate Semantic Web", Gartner "hot pick" for
2006
13
Digital asset management
Semantic Web portals
Ontology editors (and other tools)
Semantic Web and social networking
14
Significant Corporate Activity
50 Semantic Web press releases each month
15
Significant Government Activity

• Agencies moving beyond the "talk" phase
• primarily prototyping, but first acquisitions
  starting
• Example
• NASA is developing an enterprise data strategy
  around using existing data via Semantic Web
  integration

(A. Schain, 3/06)
16
There's a Lot Out There!
Paid ads
2,120,000 hits on "RDF filetyperdf"
13,600 hits on "ontology filetypeowl"
(March, 2006)
17
Where we are today

• Survey of 1300 OWL ontologies found by crawl
• Wang 06
• 19 ontologies with 2000 classes
• 6 ontologies with 10000 classes
• 2 ontologies with50000 classes
• CYC, NCI

Species RDFS OWL Lite OWL DL OWL Full Error
Count 587 199 149 337 3
18
Swoogle
http//swoogle.umbc.edu
19
Some "Swoogle" observations
The OWL namespace has been declared by 113,000
SWDs (8) and actually used by 108,000 (7). The
RDFS namespace enjoys more use, being declared by
677,000 (47) and used by 538,000 (37) SWDs.
OwlClass is the most used term from the OWL
namespace with 1,800.000 instantiations in
68,000 SWDs
We also noticed significant use of two OWL
equality assertions owlsameAs (280,000
assertions in 17,00 SWDs) and owlequivalentClass
(70,000 assertions in 4,300 SWDs). Their common
use may be an indication of increased ontology
alignment.
(Ebiquity blog, Sept 1, 2006)
20
The cake is evolving as well..
(Tim Berners-Lee)
(Tim Berners-Lee)
2001
2006
21
New languages underway

• SPARQL
• Query language for (distributed) RDF triple
  stores
• The SQL of the Semantic Web
• GRDDL/RDFa
• Integration of HTML world and Semantic Web
• Means for "embedding" RDF-based annotation on
  traditional Web pages
• Means for generating RDF triple stores from
  (annotated) Web pages
• RIF
• Rules interchange format
• Representing rules on the Web
• Linking rule-based systems together
• And more
• Multimedia annotation, Web-page Metadata
  annotation, Health Care and Life Science (LSID),
  Privacy

22
Next Steps
23
The Great Wall
24
Built in pieces at different times
Linked together for greater effect
25
The World Wide Web
Built in pieces at different times
Linking of "Web Islands"
Linked together for greater effect
26
Linking is power!
lt?xml version"1.0" encoding"UTF-8"?gt lt!DOCTYPE
rdfRDF lt!ENTITY feleuk.owl
"http//www.mindswap.org/ontologies/feleuk.owl"gt
lt!ENTITY owl "http//www.w3.org/2002/07/owl"gt
lt!ENTITY rdf "http//www.w3.org/1999/02/22-rdf-syn
tax-ns"gt lt!ENTITY rdfs "http//www.w3.org/2000/
01/rdf-schema"gt lt!ENTITY NCI
"http//www.ncibi.nih.gov/NCIT/NCIT.owl"gt
lt!ENTITY CYC"http//www.cyc.com/2004/06/04/cyc"gt
gt ltrdfRDF xmlbase"feleuk.owl"
xmlnsowl"owl" xmlnsrdf"rdf"
xmlnsrdfs"rdfs"
xmlnsNCI"NCI" xmlnsCYC"CYC"gt
ltowlOntology rdfabout""
rdfslabel"Feline Leukemia"
owlversionInfo"Feline Leuk 1.0"/gt ltowlClass
rdfabout"Feline-Leukemia"gt
ltrdfssubClassOf rdfresource"NCILeukemia"/gt
ltrdfssubClassOfgt ltowlRestrictiongt
ltowlallValuesFrom rdfresource"CYCcat"/gt
ltowlonProperty rdfresource"NCIdiseased-org
anism"/gt lt/owlRestrictiongt
lt/rdfssubClassOfgt lt/owlClassgt lt/rdfRDFgt
Link to 45000 terms at NCI
Link to 47000 (Open)CYC terms
27
Linking is power

• Today we can find thousands of ontologies
• Available on the Web
• Linked to Web resources
• Linked to data resources
• Linked to each other
• Linked to Web 2.0-like annotations
• And billions of annotated (semi-Knowledge
  engineered) objects
• Available on the Web
• Linked to Web resources
• Linked to data resources
• Linked to each other
• Linked to the ontologies

We must link these together for great effect!!
28
A key opportunity

• Vast amounts of "semi-engineered" knowledge
• Flickr tens of millions of keyword tagged photos
• Wikipedia thousands of carefully documented
  subjects (in a hierarchy, with disambiguation, )
• Etc. etc. etc.
• With "persistent" URIs
• "tank" http//en.wikipedia.org/wiki/Tank
  (armament)
• "tank" http//en.wikipedia.org/wiki/Tank2C_Pakist
  an (small town in Pakistan)
• And anything with a URI can be linked to the
  Semantic Web!!!!!

29
For exciting linking possibilities

• Linking of Web 2.0 and Semantic Web
• Using informal KE to bootstrap "formal" KE
• Extending formal KE from Web 2.0

30
Evolving vision

• Documents, linked to
• Images, annotated with
• Ontologies, linked to
• Other ontologies, describing
• Databases, exported as
• RDF graphs, as input to
• Services, which designate
• Documents, linked to
• (ad infinitum)

Stay tuned
2001
2000
1994
31
Semantic Web Challenges

• Today's Semantic Web Languages
• Are not-very-expressive-KR-language standards
• Not KIF, or even KL-ONE
• Create non-persistent knowledge bases
• Servers come and go
• Ontologies change over time
• And can't be kept consistent
• Disagreement, error, dishonesty

32
Semantic Web opportunities

• Today's Semantic Web Languages
• Are not-very-expressive-KR-language standards
• Like HTML is to SGML
• Create non-persistent KBs
• Like the 404 error (w/o which there is no Web)
• And can't be kept consistent
• Like blog-space and Web 2.0
• We need to accept, and more importantly exploit,
  these features

33
Note to Grad students (and their advisors)

• The Semantic Web today, esp at the ontology
  layer, is like the Web with no one using lta
  hrefgt
• What makes the Web, the Web
• Please, No more one ontology, one domain, one set
  of services, one Theses
• There's a reason we built this stuff on top of
  RDF and URIs

The network effect is where the power is!
34
A few of the many things I've left out

• Semantic Web Services
• Crucial for linking "programs" into the mix
• Semantic Web tools and scaling issues
• Engineering approaches being used to scale
  Semantic Web stores to database sizes
• Information extraction and Semantics
• Can we "retrofit" semantics on the existing Web
• Semantic Web Information Creations
• Can we make it so we don't have to retrofit
  future Web?
• Other information resources
• Personal data, unstructured resources, off-line
  collection information, digital libraries,
• There's more that isn't on the Web than is on it!
• New Web use patterns
• Social networks, blogs, wikis,
• are all fertile areas for Semantic Web
  exploration

35
Conclusion

• The Semantic Web is real
• Tremendous progress in the past five years
• Lots of it is out there
• Growing support in industry and govt use
• Development continues
• Easy to get involved
• Many open source tools
• New languages and techniques reaching critical
  mass
• The next steps are exciting
• The "network effect" of linking to other Semantic
  Web resources
• and to non-Semantic Web resources
• And research opportunities still abound
• Scaling
• Inconsistency
• Access and acquisition

36
(No Transcript)