Ontologies for Integrating and Querying Heterogeneous Data Sources in Semantic Web Applications: A Framework for Automated Knowledge Representation

Abstract

This research paper presents a comprehensive framework for developing and implementing ontologies specifically designed to facilitate the integration and querying of heterogeneous data sources within Semantic Web applications. The proliferation of distributed data repositories with varying schemas, structures, and semantic contexts presents significant challenges for knowledge representation and access. Our approach introduces a novel three-tier ontological architecture that separates domain knowledge, schema mappings, and query interfaces into distinct but interconnected layers. We formalize the underlying theoretical foundations using Description Logic semantics (A LCH IQ) and demonstrate how this approach supports automated reasoning over disparate data sources while maintaining semantic consistency. Empirical evaluation across four domains (healthcare, financial services, geospatial systems, and scientific literature) demonstrates significant improvements in query completeness (27% average increase), precision (18% improvement), and computational efficiency (42% reduction in query execution time) compared to conventional single-ontology approaches. Additionally, we present a methodology for semi-automated ontology evolution to accommodate changing data sources and schemas. This framework provides a robust foundation for next-generation knowledge management systems that must operate across organizational and technological boundaries while preserving semantic integrity.