Semantic Integration in Web-Based Distributed Database Systems

摘要

The ability to neaningfully integrate diverse Internet data sources is a key requirement of Web-based information systems. The processing of global queries in extranet Web distributed databases necessitates a considerable amount of intelligence and consistency resolution in determining the relevance of data items scattered across different Web databases which were typically built with disparate intranet objectives. A paradigm for he semantic integration of Web databases is presented, which makes use of an Extranet Schema of self-depicting data, and eliminates the need for a global conceptual schema. In this paradigm, there are three stages of Web data query processing: metadata access, multi-site query evaluation, and global data consolidation. The intelligence in the present structure is located at four places: (i) the human user, (ii) the data dictionary, (iii) the thesaurus, and (iv) the query coordinator. This structure enables the removal of a substantial amount of complexity from the system, which at the same time ensures a high degree of accuracy in the final query result.