Searching and Ranking Ontologies on the Semantic Web Edward Thomas (Aberdeen) Harith Alani (Southampton) Derek Sleeman (Aberdeen) Christopher Brewster.

Slides:



Advertisements
Similar presentations
Ontology-Based Computing Kenneth Baclawski Northeastern University and Jarg.
Advertisements

Fatma Y. ELDRESI Fatma Y. ELDRESI ( MPhil ) Systems Analysis / Programming Specialist, AGOCO Part time lecturer in University of Garyounis,
OMV Ontology Metadata Vocabulary April 10, 2008 Peter Haase.
Dr. Leo Obrst MITRE Information Semantics Information Discovery & Understanding Command & Control Center February 6, 2014February 6, 2014February 6, 2014.
…to Ontology Repositories Mathieu dAquin Knowledge Media Institute, The Open University From…
University of Southampton & Key Perspectives Interoperable Repository Statistics IRS irs.eprints.or g Flexible, useful, insightful and interoperable statistics.
Oyster, Edinburgh, May 2006 AIFB OYSTER - Sharing and Re-using Ontologies in a Peer-to-Peer Community Raul Palma 2, Peter Haase 1 1) Institute AIFB, University.
DC2001, Tokyo DCMI Registry : Background and demonstration DC2001 Tokyo October 2001 Rachel Heery, UKOLN, University of Bath Harry Wagner, OCLC
IST Humboldt University Berlin, Germany – Computer and Media Service – Electronic Publishing Group Birgit Matthaei, 4th Sept. 2003, Bath,
An Approach to Cope with Ontology Changes for Ontology-based Applications Yaozhong LIANG, Harith ALANI, David DUPPLAW, Nigel SHADBOLT {y.david.liang |
Supporting Complex Design using AKT technology: Rolls-Royce Case Studies Derek Sleeman: Aberdeen David Fowler: Aberdeen Gary Wills: Southampton.
Copyright 2006 Digital Enterprise Research Institute. All rights reserved. MarcOnt Initiative Tools for collaborative ontology development.
Semantic Web based Collaborative Knowledge Management LSL, ECS Feng (Barry) Tao A generic SOA for managing semantics driven domain knowledge.
Using Several Ontologies for Describing Audio-Visual Documents: A Case Study in the Medical Domain Sunday 29 th of May, 2005 Antoine Isaac 1 & Raphaël.
26/10/2008 SWESE'08 1 Enhanced Semantic Access to Software Artefacts Danica Damljanović and Kalina Bontcheva.
Alexandra Cristea & Matthew Yau 1.
PRODUCT MODELLING. Eastman C (1999). Building Product Models, CRC Press, Boca Raton Smithers T (1989). AI-based design versus geometry-based design or.
CH-4 Ontologies, Querying and Data Integration. Introduction to RDF(S) RDF stands for Resource Description Framework. RDF is a standard for describing.
A Stepwise Modeling Approach for Individual Media Semantics Annett Mitschick, Klaus Meißner TU Dresden, Department of Computer Science, Multimedia Technology.
The Semantic Web – WEEK 4: RDF
Logics for Data and Knowledge Representation Projects and thesis introduction.
27 January Semantically Coordinated E-Market Semantic Web Term Project Prepared by Melike Şah 27 January 2005.
RDF Tutorial.
Semantic Matching of candidates’ profile with job data from Linkedln PRESENTED BY: TING XIAO SARABPREET KAUR DHILLON.
Semantic Web Workshop Exploiting Synergy Between Ontologies and Recommender Systems Stuart E. Middleton, Harith Alani Nigel R. Shadbolt, David.
Hermes: News Personalization Using Semantic Web Technologies
Information Retrieval in Practice
CSCI 572 Project Presentation Mohsen Taheriyan Semantic Search on FOAF profiles.
March 17, 2008SAC WT Hermes: a Semantic Web-Based News Decision Support System* Flavius Frasincar Erasmus University Rotterdam.
Lecture Fourteen Methodology - Conceptual Database Design
ReQuest (Validating Semantic Searches) Norman Piedade de Noronha 16 th July, 2004.
Xiaomeng Su & Jon Atle Gulla Dept. of Computer and Information Science Norwegian University of Science and Technology Trondheim Norway June 2004 Semantic.
Methodology Conceptual Database Design
Overview of Search Engines
Deploying Trust Policies on the Semantic Web Brian Matthews and Theo Dimitrakos.
Provenance Metadata for Shared Product Model Databases Etiel Petrinja, Vlado Stankovski & Žiga Turk University of Ljubljana Faculty of Civil and Geodetic.
Development of Front End Tools for Semantic Grid Services Dr.S.Thamarai Selvi, Professor & Head, Dept. of Information Technology, Madras Institute of Technology,
Chapter 2 Architecture of a Search Engine. Search Engine Architecture n A software architecture consists of software components, the interfaces provided.
BLAST: A Case Study Lecture 25. BLAST: Introduction The Basic Local Alignment Search Tool, BLAST, is a fast approach to finding similar strings of characters.
NLP And The Semantic Web Dainis Kiusals COMS E6125 Spring 2010.
© Paul Buitelaar – November 2007, Busan, South-Korea Evaluating Ontology Search Towards Benchmarking in Ontology Search Paul Buitelaar, Thomas.
IPAS project: Providing a Knowledge Desktop Gary Wills, Richard Crowder, Nigel Shadbolt and Sylvia Wong July2008.
RCDL Conference, Petrozavodsk, Russia Context-Based Retrieval in Digital Libraries: Approach and Technological Framework Kurt Sandkuhl, Alexander Smirnov,
19/10/20151 Semantic WEB Scientific Data Integration Vladimir Serebryakov Computing Centre of the Russian Academy of Science Proposal: SkTech.RC/IT/Madnick.
Samad Paydar Web Technology Lab. Ferdowsi University of Mashhad 10 th August 2011.
29-30 October, 2006, Estonia 1 IST4Balt Information analysis using social bookmarking and other tools IST4Balt Information analysis using social bookmarking.
Course grading Project: 75% Broken into several incremental deliverables Paper appraisal/evaluation/project tool evaluation in earlier May: 25%
2007. Software Engineering Laboratory, School of Computer Science S E Web-Harvest Web-Harvest: Open Source Web Data Extraction tool 이재정 Software Engineering.
Q2Semantic: A Lightweight Keyword Interface to Semantic Search Haofen Wang 1, Kang Zhang 1, Qiaoling Liu 1, Thanh Tran 2, and Yong Yu 1 1 Apex Lab, Shanghai.
Problems in Semantic Search Krishnamurthy Viswanathan and Varish Mulwad {krishna3, varish1} AT umbc DOT edu 1.
Using Several Ontologies for Describing Audio-Visual Documents: A Case Study in the Medical Domain Sunday 29 th of May, 2005 Antoine Isaac 1 & Raphaël.
User Profiling using Semantic Web Group members: Ashwin Somaiah Asha Stephen Charlie Sudharshan Reddy.
Ontology based e-Real Estate Agency Information System By Moein Mehrolhasani Bijan Zamanian cmpe 588.
THE SEMANTIC WEB By Conrad Williams. Contents  What is the Semantic Web?  Technologies  XML  RDF  OWL  Implementations  Social Networking  Scholarly.
1 Open Ontology Repository initiative - Planning Meeting - Thu Co-conveners: PeterYim, LeoObrst & MikeDean ref.:
Jens Hartmann York Sure Raphael Volz Rudi Studer The OntoWeb Portal.
Characterizing Knowledge on the Semantic Web with Watson Mathieu d’Aquin, Claudio Baldassarre, Laurian Gridinoc, Sofia Angeletou, Marta Sabou, Enrico Motta.
OWL Web Ontology Language Summary IHan HSIAO (Sharon)
Supporting Collaborative Ontology Development in Protégé International Semantic Web Conference 2008 Tania Tudorache, Natalya F. Noy, Mark A. Musen Stanford.
Selected Semantic Web UMBC CoBrA – Context Broker Architecture  Using OWL to define ontologies for context modeling and reasoning  Taking.
General Architecture of Retrieval Systems 1Adrienn Skrop.
26/02/ WSMO – UDDI Semantics Review Taxonomies and Value Sets Discussion Paper Max Voskob – February 2004 UDDI Spec TC V4 Requirements.
Dmitry Mouromtsev, Aleksei Romanov, Dmitry Volchek and Fedor Kozlov Laboratory ITMO University, St. Petersburg, Russia “Metadata Extraction from.
ONTOLOGY LIBRARIES: A STUDY FROM ONTOFIER AND ONTOLOGIST PERSPECTIVES Debashis Naskar 1 and Biswanath Dutta 2 DSIC, Universitat Politècnica de València.
OWL (Ontology Web Language and Applications) Maw-Sheng Horng Department of Mathematics and Information Education National Taipei University of Education.
Information Retrieval in Practice
Exploiting Synergy Between Ontologies and Recommender Systems
SIS: A system for Personal Information Retrieval and Re-Use
Ontology Evaluation ارزیابی آنتولوژی
Introduction to Information Retrieval
Presentation transcript:

Searching and Ranking Ontologies on the Semantic Web Edward Thomas (Aberdeen) Harith Alani (Southampton) Derek Sleeman (Aberdeen) Christopher Brewster (Sheffield) KCAP-05 Workshop Ontology Management: Searching, Selection, Ranking, and Segmentation Third International Conference on Knowledge Capture Banff, Canada

Other Ontology Search engines Google offers a powerful search engine; but for ontological information, does not provide good visualization & does not allow the user to make some important distinctions (see later) Google offers a powerful search engine; but for ontological information, does not provide good visualization & does not allow the user to make some important distinctions (see later) Swoogle allows one to search for classes or properties, but does not allow one to search for structural features (eg animal class with cat within 2 links) Swoogle allows one to search for classes or properties, but does not allow one to search for structural features (eg animal class with cat within 2 links)

Specification for OntoSearch We implemented a basis system with Google as its heart & then had users evaluate it. Heres a summary of their requirements: We implemented a basis system with Google as its heart & then had users evaluate it. Heres a summary of their requirements: –The ability to specify the types of file(s) to be returned (OWL, RDFS, all) –The ability to specify the types of entities to be matched by each keyword (concept, attribute, values, comments, all) –The ability to specify partial or exact matches on entities. –Sub-graph matching eg concept animal with concept pig within 3 links; concepts with particular attributes

Implementation The heart of the enhanced system is a repository of previously run Queries and responses provided by Google; this info is date stamped The heart of the enhanced system is a repository of previously run Queries and responses provided by Google; this info is date stamped (Non-structural) User queries are broken down into keywords & Google queries are constructed for these keywords (Non-structural) User queries are broken down into keywords & Google queries are constructed for these keywords Before each query is run, the Repository is checked to see if that query has been run less than D days ago; Before each query is run, the Repository is checked to see if that query has been run less than D days ago; If the query is a new one or the query was run > D days ago then the query is run on Google & the results are indexed as above. If the query is a new one or the query was run > D days ago then the query is run on Google & the results are indexed as above. Then the original user query is answered using the (updated) repository Then the original user query is answered using the (updated) repository Finally, Results are returned to the user Finally, Results are returned to the user OntoSearch is implemented in Java Servlets OntoSearch is implemented in Java Servlets [Repository uses Berkeley DB2 with optimised indexes for better performance than SQL based Triple Stores] [Repository uses Berkeley DB2 with optimised indexes for better performance than SQL based Triple Stores]

OntoSearch OntoSearch currently has two main interfaces OntoSearch currently has two main interfaces –Keyword based class match search Simple interface allows matching on class and property names Simple interface allows matching on class and property names –A query based structure search

Example Query Query syntax is similar to the N/Triples representation: This Query contains four conditions which must be met in a matching document: contains an element $1 which has type #Ontology contains an element $1 which has title Ontology contains an element $2 which has type #Class contains an element $2 which has a comment containing programmer Keywords (Ontology, programmer) are extracted from this query and used to query Google in the same way as would be done for the standard search, and any results which come back are downloaded and added to the repository.

Ontology Ranking Current ranking systems (Swoogle, OntoKhoj) rely on an ontologys popularity to determine its rank Current ranking systems (Swoogle, OntoKhoj) rely on an ontologys popularity to determine its rank –The large number of FOAF and RSS files on the Semantic Web produce a large bias in favour of ontologies used by these representations –The relatively small number of other Semantic Web Documents does not constitute a large enough data set to make reliable judgements No consideration is given as to how a matching concept is represented in an ontology No consideration is given as to how a matching concept is represented in an ontology –Searching for an ontology to describe a student may match a well connected ontology which covers the whole academic domain –The popularity of a general ontology is likely to be higher than a specific ontology which better covers the particular concepts searched for

AKTive Rank AKTive Rank uses four measures to determine an Ontologys rank (detail given in the paper) AKTive Rank uses four measures to determine an Ontologys rank (detail given in the paper) –Class Match Measure –Centrality Measure –Density Measure –Semantic Similarity Measure These measures are weighted and summed to give an overall AKTive Rank score These measures are weighted and summed to give an overall AKTive Rank score

Example OntoSearch was queried in class match mode for ontologies matching the concept student and university OntoSearch was queried in class match mode for ontologies matching the concept student and university 8 ontologies were returned, 2 failed basic tests; so 6 were analysed by the AKTive Rank system 8 ontologies were returned, 2 failed basic tests; so 6 were analysed by the AKTive Rank system –dan - –ita - –univ-bench - –swportal - akt_ontology_LITE_inst.owl –ka - –russia2 - karlsruhe.de/WBS/meh/foam/ontologies/russia2.owl Weighted sums were calculated for each ontology, and the sums ranked Weighted sums were calculated for each ontology, and the sums ranked

Results Rank AKTiveRank Score SSM DEM CMM CEM sw- portal kaitadan Russia 2 univ- bench WeightingNameOntologiesMeasure

Future Work OntoSearch OntoSearch –More advanced query language will be implemented –Additional Web Services / APIs –Additional sources of ontologies (including private ontologies) will need to be integrated –Better integration with AKTive Rank –Additional visualisation tools are being developed AKTive Rank AKTive Rank –The parameters used in the AKTive Rank process need to be reconsidered following further evaluations by knowledge Engineers of the ontologies recommended. –Existing RDF query languages are inadequate for dealing with graph queries; incorporate a better graph querying system, such as JUNG.

Thank You Any Questions? Work done as part of the AKT consortium Christopher Brewster Computer Science Dept. Uni. of Sheffield, Sheffield, UK Derek Sleeman Computer Science Dept. Aberdeen University Aberdeen, UK Harith Alani Dept. of Electronics and Computer Science Uni. of Southampton Southampton, UK Edward Thomas Computer Science Dept. Aberdeen University Aberdeen, UK