An OO Model for Incremental Hierarchical KA
Using a database management system (DBS) to build a knowledge base system (KBS) is sometimes desirable because DBS systems allow management of large sets of rules, control of concurrent access and managing multiple knowledge bases simultaneously. In this paper, we describe how to build a KBS using a database management system DBMS for its schema evolution ability. This allows the use of an Object Oriented DMBS (OODBMS) to manage the consistency of an incrementally built hierarchical knowledge base (KB). The underlying knowledge representation scheme, which we use, is our Nested Ripple Down Rules (NRDR).
An NRDR KB evolves into a hierarchy of concepts where each concept is defined as a collection of hierarchical rules with exceptions. To modify a concept definition, exception rules are added by a domain expert, they are never deleted or modified. This eases maintenance and development of a concept definition, but may cause inconsistencies to occur in the KB. We analyse the relation between cases and rules as the knowledge base evolves and as these inconsistencies occur. We explore what specific features an OO database model should accommodate to be used to implement an NRDR KB. The aim is that this in turn allows the use of the built-in mechanisms to manage the consistency of an evolving NRDR conceptual hierarchy. The significance of this paper is two folds: first, it describes an efficient mechanism maintaining consistency of an evolving classification hierarchy, using built-in schema evolution features of an OODBMS. Second, it proposes an intelligent interface for an OODBMS, to allow intelligent classification queries over stored objects.
Unable to display preview. Download preview PDF.
- 1.Al-Jadir, L. and G. Beydoun. Using F2 OODBMS to Support Incremental Knowledge Acquisition. in International Database Engineering & Applications Symposium (IDEAS02). 2002. Canada.Google Scholar
- 2.Al-Jadir, L., et al. Evolution Features of the F2 OODBMS in 4th International Conference on Database Systems for Advanced Applications (DASFAA95). 1995. Singapore: World Scientific.Google Scholar
- 3.Beydoun, G. and A. Hoffmann. Acquisition of Search Knowledge, in The 10th European Knowledge Acquisition Workshop (EKAW97). 1997. Spain: Springer.Google Scholar
- 4.Beydoun, G. and A. Hoffmann. A Holistic Approach for Knowledge Acquisition in 11th European Conference on Knowledge Acquisition and Management (EKAW99). 1999. Germany: Springer.Google Scholar
- 5.Beydoun, G. and A. Hoffmann. Monitoring Knowledge Acquisition, Instead of Evaluating Knowledge Bases. in 12th European Conference on Knowledge Acquisition and Knowledge Management (EKAW2000). 2000. France: Springer.Google Scholar
- 7.Clancey, W.J., The Knowledge Level Reinterpreted: Modelling How Systems Interact. Machine Learning, 1989. 4: p. 285–291.Google Scholar
- 8.Compton, P. and R. Jansen. Knowledge in Context: a strategy for expert system maintenance. in Second Australian Joint Artificial Intelligence Conference (AI88). 1988.Google Scholar
- 9.Richards, D. and P. Compton. Knowledge Acquisition First, Modelling Later. in 10th European Knowledge Acquisition Workshop (EKAW97). 1997. Spain: Springer.Google Scholar
- 10.Tamma, V.A.M. and T.J.M. Bench-Capon. Supporting Inheritance Mechanisms in Ontology Representation, in 12th International Conference on Knowledge Engineering and Knowledge Management (EKAW2000). 2000. France: Springer.Google Scholar
- 11.Wiratunga, N. and S. Craw. Informed Selection of Training Examples for Knowledge Refinement, in 12th International Conference on Knowledge Engineering and Knowledge Management. 2000. France: Springer.Google Scholar