Ontologies are not the Panacea in Data Integration: A FlexibleCoordinator to Mediate Context Construction

  • Authors:

  • Aris M. Ouksel;Iqbal Ahmed

  • Affiliations:

  • The University of Illinois at Chicago, Department of Information and Decision Sciences (M/C 294), Chicago, IL 60607. E-mail: aris@uic.edu;The University of Illinois at Chicago, Department of Information and Decision Sciences (M/C 294), Chicago, IL 60607. E-mail: iqahmed@uic.edu

  • Venue:
  • Distributed and Parallel Databases
  • Year:
  • 1999

Quantified Score

Hi-index 0.00

Visualization

Abstract

Shared ontologies describe concepts and relationships to resolve semantic conflicts amongst users accessing multiple autonomous andheterogeneous information sources. We contend that while ontologies areuseful in semantic reconciliation, they do not guarantee correctclassification of semantic conflicts, nor do they provide the capability tohandle evolving semantics or a mechanism to support a dynamic reconciliationprocess. Their limitations are illustrated through a conceptual analysis ofseveral prominent examples used in heterogeneous database systems and innatural language processing. We view semantic reconciliation as anonmonotonic query-dependent process that requires flexible interpretationof query context, and as a mechanism to coordinate knowledge elicitationwhile constructing the query context. We propose a system that is based onthese characteristics, namely the SCOPES (Semantic Coordinator Over ParallelExploration Spaces) system. SCOPES takes advantage of ontologies to constrainexploration of a remote database during the incremental discovery andrefinement of the context within which a query can be answered. It uses anAssumption-based Truth Maintenance System (ATMS) to manage the multipleplausible contexts which coexist while the semantic reconciliation process isunfolding, and the Dempster-Shafer (DS) theory of belief to model thelikelihood of these plausible contexts.