Ontology-based approach for gathering the heterogeneous information sources

Abstract

The problem of accessing and integrating heterogeneous information sources has become the center stage in the distributed processing environment. One of the important issues stemming from accessing these heterogeneous sources is semantic heterogeneity. A number of systems have been proposed to address this issue, ranging from mediator based systems to description logic-based systems. However, the current methodologies for accessing heterogeneous sources offer limited flexibility, scalability, interoperability, and robustness. This dissertation proposes the Semantic Information Gathering Approach, hereafter SIGA, a system for accessing and integrating heterogeneous sources on the WWW. The reference architecture is based on layered-architecture incorporating mobile agents as the client/server connectivity to eliminate traditional client/server overhead. A metadata dictionary which is an essential component of the reference architecture is also proposed for solving semantic heterogeneity. The metadata dictionary is derived from domain ontology where the constituent components are defined in terms of object oriented principle and set theory. To enable system-wide interoperability suitable for a Web-based environment, the XM L technology is selected as the language for expressing the metadata dictionary contents. Consequently, this dissertation also covers the overall system and agent architecture, the modeling process of the ontology-based metadata dictionary, the components and representation of the metadata dictionary, and the querying process in accessing and integrating the heterogeneous sources through the metadata dictionary.