Ontology Driven Query Expansion

Abstract

Today‟s information saturated world produces more data than ever before. It is very difficult for the users, scientists and domain experts to query and analyze the desired data from autonomous and geographically distributed heterogeneous data sources, as most of them are naïve and do not have knowledge of internal structure of the data sources, their contents and query languages. Users require the uniform and transparent access to the heterogeneous data sources. Most of the data integration architectures provide uniform access to the underlying data sources usually by employing the mediator-wrappers architecture. Semantic queries help in transparent access. They are generated with the support of semantic model (domain ontology) by making use of relationships between concepts. This research focuses on expanding users queries to provide transparent access and improvements in recall, so that information loss is minimized. We propose a solution for expanding queries to extend the results of user query by using the ontology. We have developed two rules for expanding user queries semantically by extending the previous work. We have implemented both the previous and proposed techniques independently and checked the results of queries for both techniques. It is evident from the evaluation of results that recall is improved in our proposed technique with respect to the previous technique. Hence information loss is minimized.