Algorithms for query processing in a distributed knowledge integration system

Abstract

The main goal of the paper is to introduce a framework of knowledge integration system. The framework is based on the ontology integration system described by Calvanese in [1], [2] and extends it to a multi-layer framework. The paper also shows how usage of cartographic representation of knowledge [3], [4] to store and process ontologies in knowledge sources and building mappings between ontologies allows for implementing efficient algorithms of query processing in a distributed environment of a knowledge integration system. A new method of creating mappings between different terminologies, called CMC, is described. Using the cartographic representation it is possible to map regions (that correspond to all satisfiable logical combinations of terms) defined in one ontology to regions of another one. Moreover, SPIDER - a query processing algorithm, based on cartographic mappings, is described. The paper ends with a presentation of the complexity of presented methods and algorithms.