The Semantic Web Challenges with today s Web the Semantic Web Technology Example of use Status Semantic Web in e-learning 1
The Web - the worlds largest library? Not catalogued and categorized No professional staff
The Web 2.0 Two major efforts to advance the Web Web Services Semantic Web
Web Services Open up legacy catalogs and systems to the Web Deep information Structured information (but buried in legacy systems) Goal: Interoperability
The Semantic Web Open up human readable assets for machines Shallow information Unstructured (but linked through the web) Goal: Enhance search (short term) and enable automated deduction (long term)
Trends in information management The Web & unstructured information (blogs, search, wiki, visualization) Web Services & structured information (catalogues, b2b integration, web services) Semantics Micro management & aggregation (Micro payment, DRM, RFID, sensor networks, RSS) Personalized information (instant messaging, e-mail, collaboration)
Evolution of the Web today The shallow web The deep web HTML enables sharing of documents on the Web Databases & legacy systems gets enabled by Web access technologies (CGI, Perl, Java,.Net) Search engines enable humans to reach into documents and later catalogs and databases Catalogs and services opens up with Web Services and XML Metadata and URI s enabling resources to be described uniquely, enabling smarter search 1990 s 2000 s 2005-2010- Business processes opens up with Web Services Ontologies and rules enables deduction over the Web Everything gets Web enabled (RFID, sensor networks etc) Web 2.0 Services and resources merges time and gets enabled through The full vision of The semantic web
The semantic Web
Semantic Web Principles Anything can have a URI#xxx. Vocabularies can merge and be replaced with time Documents are selfdescribing "Anyone can say anything about anything" No one system knows everything Design must be minimalist
Mathematics & Logic Semantics Knowledge foundation
Where are we today? XML and Unicode (well established standards) Metadata description language (RDF 1.0 due Dec. 2003) Annotations and Schema (RDF Schema, RSS, Dublin Core)
Data format architecture Web Services, ebxml, RosettaNet, MathML, xhtml, WML etc. 1990 2000 2010
Evolution of web technologies RDF Machine understandable documents XML Machine readable documents and services HTML Human readable and understandable documents with additional standards like UTF-8 including Web Services Technologies including RDFS and OWL technologies 13
Emerging Emerging Technologies Technologies Gartner Research
Status Semantic Web W3C
Technology
Metadata and RDF/XML XML = syntax RDF = structure Metadata = semantics & resources
Layers
Interoperability RDF presents a single storage model Challenge: translating between the different formats used for VCard, Dublin Core, Dublin Core Qualifiers, IEEE LOM, SCORM etc., which are meta-data standards without the common RDF format
The triple <rdf:description rdf:about="subjectofmytriple"> <PredicateOfMyTriple>ObjectOfMyTriple</PredicateOfMyTriple> </rdf:description>
Annotation in RDF 2003-07-16T19:20:30Z a:created Photograph of lake lined by mountains. Blue sky and green and yellow fields also visible. David Norheim dc:description foaf:name http://www.asemantics.com/picture.jpg dc:creator rdf:resource dc:language dc:title dc:subject Lake and mountain view http://foaf.asemantics.com/david.rdf en Lake3
METADATA specifications Dublin Core IMS ARIADNE IEEE LTSC LOM ADL SCORM ISO 11179 AICC ALIC (Japan) CEN Cisco RLO MARC 21 (MARBI & OAI) Msoft LRN Warwick Framework Z39.5 EML
1. Everything has a URI Don't say "color" say "http://www.pantomime.com/2002/std6#color" In RDF/XML <rdf:rdf xmlns:pan= http://www.pantomime.com/2002/std6# > <rdf:description rdf:about="#sky"> <pan:color>blue</pan:color> </rdf:description> </rdf:rdf>
What s ahead Tools and apps - just as the web is a killer app for the internet the semantic web is a killer app at that magnitude Where is the money? Ontologies for vertical markets Data integration US Government, access Enterprise application market, business process management enabled by Web Services Composing semantically 24
What is machine processable content, it is actually one of the original design ideas of the semantic web, uses metadata to add & extract meaning, builds upon and extends existing internet infrastructure (rewrite)
A semantic cake End user application or interface classical application e.g. RDBMS + web interface application application RDFS RDF application OWL Gradually structuring data Unstructured and structured data 26
Uses
Uses Context aware links (personalized through foaf) Collaborative filtering (rating through pics 2.0) Collaborative categorization and annotation Correction (through annotation)
How do I get to Reno the fastest? Agent FOAF (foaf.asemantics.com/david) Weather (rdf.weatherchannel.com) Traffic Info (rdf.sfpd.com) Map (maps.yahoo.com) Airline info (rdf.sfo.com) an francisco weathertoday weathertomor row rainy cold Reno access Bus #31 access Bus #21 route route 101 N Union St route route 101 N Union St 29
Still a lot to be done Semantic annotation (e.g. Amaya) Ontologies (a lot of research) Lexical entities (and information extraction) Search functions (e.g. RDQL, Jena) Search engines has two major drawbacks: Coverage and Expressiveness (no semantics)
Semantic Web & e- learning
Potential for learning Intelligent agents can adjust curriculum Intelligent tutors can dynamically assemble learning objects Content can self-adjust tor be created (source: Advanced Distributed Learning)
Dublin Core What is the relationship between Dublin Core Metadata and RDF and XML? Dublin Core Metadata Element Set (IETF RFC 2413) and RDF (Resource Description Framework) (W3C Recommendation) are two distinct specifications. Neither requires the other, but their co-evolution forms a natural complement within the Web's greater metadata architecture.
And LOM? Learning Objects Metadata (LOM) Binding to RDFS and interoperable with Dublin Core
What about SCORM? Most of the work so far has focused on XML as the encoding language for such specifications (e.g. IMS, IEEE-LOM, and SCORM) SCORM [Dod02] has chosen its own XML formats and methodologies, thereby limiting the educational community to a restricted universe, and making it much more difficult to integrate e-learning with other business processes.
Semantic Web in E-learning (1) Intelligent software agents can be implemented, helping the learner to find and use globally distributed learning resources. Personal annotations of any learning resource becomes a feasible technology, as demonstrated by Annotea
Semantic Web in E-learning (2) Collaborative and distributed authoring and course construction becomes much simpler thanks to the modularity of the information. Reuse of learning material by crossfertilization suddenly becomes a reality, creating important synergy effects.
Enough theory... an example please! san francisco weathertoday rainy weathertomorrow cold North Beach access Bus #31 route 101N access Bus #21 route Union St 38
Web Services and the Semantic Web classical application Application web services wrapper Native or web service driven application existing resource Semantic Application semantic wrapping or annotation Semantic driven application using ontologies 39
Web of Semantics A little semantics Goes a long way! nn, keynote speaker This create a powerful network effect! We won t need the reasoning, models = how you communicate among groups to create an agreement (water molycuse - the model is clear...), abstraction and links... An old idea in New clothes graph (rdf), graph+logic (owl) Network effekt = much more powerful, through abstraction we can share concepts. small amount can be powerful Challenge: How do exploit the web of semantics Cows in texas = I just happen to be on the side of the majority at the moment... Personal information created the old web 40
Challenges Web Services is ready for prime time Process modeling needs work (runtime, not for analyzing) Ontologies, what is right? 41
Challenges End user tools Power forms from ontologies 1) The hammer, etc (RDF) - most people would need this - interoperability 2) power tools (ontologies = knoledge representation) - what has the web told us?... nominal view on the world. (nouns)..no processes - that s web services (verbs) => languages? 3) AI and the Semantic Web further ahead - agents, reality: shared but only partical, does not share common ontologies... AI: Knowledge engineering -pure PS: look at Tim Beerners lee s one pages articles on w3c e.g. small world network... 42
Who is behind? Who is behind? W3C HP Sun IBM Open Source Applications Foundation (OSAF), MovableType.org, Teknowledge Creative Commons.org Netsemantics Applied Minds Asemantics +++++++
Thank you david.norheim@ntc.no