Background and Motivation Contributions Multi-Perspective Ontology Engineering R. Denaux 1 A. G. Cohn 1 V. Dimitrova 1 G. Hart 2 1 School of Computing University of Leeds 2 Ordnance Survey Research Invited Talk at Department of Computer Science Sheffield, 2010 Denaux, Cohn, Dimitrova and Hart Multi-Perspective Ontology Engineering
Description of my PhD for an invited talk at the Department of Computer Science at Sheffield University. I present my hypothesis that different ontology authors have different views on the world, and different requirements of ontologies which results in different definitions of 'the same' concepts.
I propose to make ontology tools that are bias-aware to help in the creation and reuse of ontologies. In this context I present my current work on making a goal-aware ontology editor.
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Background and MotivationContributions
Multi-Perspective Ontology Engineering
R. Denaux1 A. G. Cohn1 V. Dimitrova1 G. Hart2
1School of ComputingUniversity of Leeds
2Ordnance Survey Research
Invited Talk at Department of Computer ScienceSheffield, 2010
Denaux, Cohn, Dimitrova and Hart Multi-Perspective Ontology Engineering
Background and MotivationContributions
Outline
1 Background and MotivationMulti-perspective Ontology EngineeringOntology Purposes
2 ContributionsGoal-aware Ontology EditorOntology Purpose VocabularyFormalisationCurrent Work
Denaux, Cohn, Dimitrova and Hart Multi-Perspective Ontology Engineering
ROO: Rabbit to OWL Ontology Authoring.Example of adapting to ontology contributors
Domain experts:Good knowledge of the domain to be representedLimited or no Ontology Engineering experienceLimited or no knowledge of OWL, Protégé, etc.
ROO provides tool support for domain experts:Guidance through ontology construction methodologyControlled Natural Language interfaceNo OWL specific terminology
Adaptation at design time, not at runtimeRe-use techniques from User Modelling andPersonalisation
Denaux, Cohn, Dimitrova and Hart Multi-Perspective Ontology Engineering
Ontology Development 101Natalya F. Noy and Deborah L. McGuinness
There is no one correct way to model a domain thereare always viable alternatives. The best solutionalmost always depends on the application that youhave in mind and the extensions that you anticipate.
. . . deciding what we are going to use the ontology for
. . . will guide many of the modeling decisions down theroad.1
1N. F. Noy and D. Mcguinness. Ontology development 101: A guide tocreating your first ontology, 2000
Denaux, Cohn, Dimitrova and Hart Multi-Perspective Ontology Engineering
The goal of the specification phase is to produceeither an informal, semi-formal or formal ontologyspecification document. METHONTOLOGY proposesthat at least the following information be included:(a) The purpose of the ontology, including its
intended uses, scenarios of use, end-users, etc.. . . 2
2M. Fernandez-Lopez, A. Gomez-Perez, and N. Juristo. Methontology:from ontological art towards ontological engineering. In Proceedings of theAAAI97 Spring Symposium Series on Ontological Engineering, pages 33–40,1997
Denaux, Cohn, Dimitrova and Hart Multi-Perspective Ontology Engineering
Local adaptation: once the core ontology is available,users work with it and adapt it locally to their ownneeds. Typically, they will have their own businessrequirements and correspondingly change their localontologies. 3
3Denny Vrandecic, H. Sofia Pinto, York Sure, and Christoph Tempich. Thediligent knowledge processes. Journal of Knowledge Management,9(5):85–96, OCT 2005
Denaux, Cohn, Dimitrova and Hart Multi-Perspective Ontology Engineering
Background and MotivationContributions
Goal-aware Ontology EditorOntology Purpose VocabularyFormalisationCurrent Work
Outline
1 Background and MotivationMulti-perspective Ontology EngineeringOntology Purposes
2 ContributionsGoal-aware Ontology EditorOntology Purpose VocabularyFormalisationCurrent Work
Denaux, Cohn, Dimitrova and Hart Multi-Perspective Ontology Engineering
Background and MotivationContributions
Goal-aware Ontology EditorOntology Purpose VocabularyFormalisationCurrent Work
Purpose-driven Adaptive Ontology Reuse.
Derive model of Ontology PurposesUse this model to capture new ontology purposesSuggest ontology to reuse based on matching ontologypurpose
Denaux, Cohn, Dimitrova and Hart Multi-Perspective Ontology Engineering
Background and MotivationContributions
Goal-aware Ontology EditorOntology Purpose VocabularyFormalisationCurrent Work
Denaux, Cohn, Dimitrova and Hart Multi-Perspective Ontology Engineering
Background and MotivationContributions
Goal-aware Ontology EditorOntology Purpose VocabularyFormalisationCurrent Work
Outline
1 Background and MotivationMulti-perspective Ontology EngineeringOntology Purposes
2 ContributionsGoal-aware Ontology EditorOntology Purpose VocabularyFormalisationCurrent Work
Denaux, Cohn, Dimitrova and Hart Multi-Perspective Ontology Engineering
Background and MotivationContributions
Goal-aware Ontology EditorOntology Purpose VocabularyFormalisationCurrent Work
Purpose Example
ontology: Ordnance Survey Hydrology Ontology v2source: ontology annotationfree text: "Purpose: To describe in an unambiguousmanner the inland hydrology feature classes surveyed byOrdnance Survey with the intention of improving the use ofthe surveyed data by our customers and enablingsemi-automatic processing of these data."
Denaux, Cohn, Dimitrova and Hart Multi-Perspective Ontology Engineering
Background and MotivationContributions
Goal-aware Ontology EditorOntology Purpose VocabularyFormalisationCurrent Work
Representing Ontology Purposes.Deriving Vocabulary from a corpus
44Ronald Denaux, Anthony G. Cohn, Vania Dimitrova, and Glen Hart.
Towards modelling the intended purpose of ontologies: A case study ingeography. In Proceedings of the Terra Cognita Workshop, collocated withthe 8th International Semantic Web Conference (ISWC-2009), volume 518.CEUR-WS, 2009
Denaux, Cohn, Dimitrova and Hart Multi-Perspective Ontology Engineering
Background and MotivationContributions
Goal-aware Ontology EditorOntology Purpose VocabularyFormalisationCurrent Work
Purpose Categories
Domain DefiningOntology Process RelatedData Process RelatedInvestigativeCollaboration EnhancingExternal ApplicationAnalogous
Example
Code Task Focus RestrictionsOS1 Describe Domain Domain is restricted
to feature classessurveyed byOrdnance Survey
Denaux, Cohn, Dimitrova and Hart Multi-Perspective Ontology Engineering
Background and MotivationContributions
Goal-aware Ontology EditorOntology Purpose VocabularyFormalisationCurrent Work
Purpose Categories
Domain DefiningOntology Process RelatedData Process RelatedInvestigativeCollaboration EnhancingExternal ApplicationAnalogous
Example
Code Task Focus RestrictionsPont5 Facilitate Ontology Process is the
Process alignment ofontologies
Denaux, Cohn, Dimitrova and Hart Multi-Perspective Ontology Engineering
Background and MotivationContributions
Goal-aware Ontology EditorOntology Purpose VocabularyFormalisationCurrent Work
Purpose Categories
Domain DefiningOntology Process RelatedData Process RelatedInvestigativeCollaboration EnhancingExternal ApplicationAnalogous
Example
Denaux, Cohn, Dimitrova and Hart Multi-Perspective Ontology Engineering
Background and MotivationContributions
Goal-aware Ontology EditorOntology Purpose VocabularyFormalisationCurrent Work
Outline
1 Background and MotivationMulti-perspective Ontology EngineeringOntology Purposes
2 ContributionsGoal-aware Ontology EditorOntology Purpose VocabularyFormalisationCurrent Work
Denaux, Cohn, Dimitrova and Hart Multi-Perspective Ontology Engineering
Background and MotivationContributions
Goal-aware Ontology EditorOntology Purpose VocabularyFormalisationCurrent Work
Representing Ontology Purposes in OWLGoals
Enable ontology contributors to formalise their ontologypurposeAllows users to express themselves in a manner that isclose to the way they would normally use (free text)classify ontologies based on their purposeCan be extended by ontology contributorsStarting point: usable but not intended to cover all possiblepurposes
Denaux, Cohn, Dimitrova and Hart Multi-Perspective Ontology Engineering
Background and MotivationContributions
Goal-aware Ontology EditorOntology Purpose VocabularyFormalisationCurrent Work
Example Formalisation
OS ExampleOS Hydrology Ontology aims to describe the OSHydrology Feature Classes.OS Hydrology Ontology intends to enable theSemi-automatic Processing of Ordnance Survey Data.OS Hydrology Ontology intends to improve the Data Usageof Ordnance Survey Customers.
Inferences we wantOS Hydrology Ontology is a Data Processing Ontology.OS Hydrology Ontology is a Domain Specifying Ontology.
Denaux, Cohn, Dimitrova and Hart Multi-Perspective Ontology Engineering
Background and MotivationContributions
Goal-aware Ontology EditorOntology Purpose VocabularyFormalisationCurrent Work
Abstract and Concrete levels
Abstract Concepts and RelationsAgent, Artifact, Creation Action, has (purpose) focus, PurposeFocus, Purpose Task.A Data Processing Ontology is anything that: is a kind ofOntology; has purpose focus at least one Data Process.
Concrete Concepts and Relationsaims to describe, describes, aims to facilitate, aims to provide,intends to enable, data process.The relationship intends to improve is a special type of therelationship has purpose focus.
Denaux, Cohn, Dimitrova and Hart Multi-Perspective Ontology Engineering
Background and MotivationContributions
Goal-aware Ontology EditorOntology Purpose VocabularyFormalisationCurrent Work
Outline
1 Background and MotivationMulti-perspective Ontology EngineeringOntology Purposes
2 ContributionsGoal-aware Ontology EditorOntology Purpose VocabularyFormalisationCurrent Work
Denaux, Cohn, Dimitrova and Hart Multi-Perspective Ontology Engineering
Background and MotivationContributions
Goal-aware Ontology EditorOntology Purpose VocabularyFormalisationCurrent Work
Purpose Elicitation Dialogue
Elicit formal description of ontology purpose in order toclassify ontology.Aid user to make transition from free text to formalrepresentation.Use Ontology of Purpose Ontologies to guide dialogueepisode.Dialogue at concrete level, reasoning at abstract level.
Denaux, Cohn, Dimitrova and Hart Multi-Perspective Ontology Engineering
Background and MotivationContributions
Goal-aware Ontology EditorOntology Purpose VocabularyFormalisationCurrent Work
Basic Dialogue Structure
Elicit free text purpose descriptionNLP analysis to generate purpose hypothesesClarify and confirm hypotheses:
elicit instances linked to concrete conceptssucceeds when classification can be made
Denaux, Cohn, Dimitrova and Hart Multi-Perspective Ontology Engineering
Background and MotivationContributions
Goal-aware Ontology EditorOntology Purpose VocabularyFormalisationCurrent Work
Basic Dialogue Structure
Elicit free text purpose descriptionNLP analysis to generate purpose hypothesesClarify and confirm hypotheses:
elicit instances linked to concrete conceptssucceeds when classification can be made
AnnotationsTo describe in an unambiguous manner the inland hydrologyfeature classes surveyed by Ordnance Survey with the intentionof improving the use of the surveyed data by our customers andenabling semi-automatic processing of these data.
Denaux, Cohn, Dimitrova and Hart Multi-Perspective Ontology Engineering
Background and MotivationContributions
Goal-aware Ontology EditorOntology Purpose VocabularyFormalisationCurrent Work
Basic Dialogue Structure
Elicit free text purpose descriptionNLP analysis to generate purpose hypothesesClarify and confirm hypotheses:
elicit instances linked to concrete conceptssucceeds when classification can be made
Hypothesis Domain Describing Ontology
OS Hydrology Ontology aims to describe X. X is a KnowledgeDomain.
Denaux, Cohn, Dimitrova and Hart Multi-Perspective Ontology Engineering
Background and MotivationContributions
Goal-aware Ontology EditorOntology Purpose VocabularyFormalisationCurrent Work
Basic Dialogue Structure
Elicit free text purpose descriptionNLP analysis to generate purpose hypothesesClarify and confirm hypotheses:
elicit instances linked to concrete conceptssucceeds when classification can be made
Hypothesis Data Processing Ontology
OS Hydrology Ontology intends to enable X. X is a Data Usage.
Denaux, Cohn, Dimitrova and Hart Multi-Perspective Ontology Engineering
Background and MotivationContributions
Goal-aware Ontology EditorOntology Purpose VocabularyFormalisationCurrent Work
Architecture
Denaux, Cohn, Dimitrova and Hart Multi-Perspective Ontology Engineering
Background and MotivationContributions
Goal-aware Ontology EditorOntology Purpose VocabularyFormalisationCurrent Work
Plan
Implement Dialogue Plan generation for a few purposetypesEvaluate Dialogue Plan generation without RabbitinterpreterAdd Rabbit interpreter and evaluate with real usersSuggest ontology to reuse based on matching ontologypurpose
Denaux, Cohn, Dimitrova and Hart Multi-Perspective Ontology Engineering
Background and MotivationContributions
Goal-aware Ontology EditorOntology Purpose VocabularyFormalisationCurrent Work
The End
Denaux, Cohn, Dimitrova and Hart Multi-Perspective Ontology Engineering
Acknowledgements
Creative Commons Images from flckr.comSurveyor image by Wessex ArcheologyWater Ecologist by lindenbaumFlood Rescuer by Tree & J HensdillOrienteerer by Tarnie
Dilbert visionary strip, copyright UFS, Inc.
Denaux, Cohn, Dimitrova and Hart Multi-Perspective Ontology Engineering