Optimized ontology-driven query expansion using map-reduce framework to facilitate federated queries

摘要

In view of the need for a highly distributed and federated architecture, a robust query expansion has great impact on the performance of information retrieval in a specific domain. We aim to determine ontology-driven query expansion terms using different weighting techniques to determine the most k-top relevant terms. For this, first we consider each individual ontology and user query keywords to determine the Basic Expansion Terms (BET). Second, we specify New Expansion Terms (NET) by Ontology Alignment (OA). Third, we use a Map-Reduce distributed algorithm for calculating all the shortest paths in ontology graph as a meta data to calculate weights for terms e BET ∪NET. Fourth, we actually weight expanded terms using a combination of semantic metrics namely Density Measure (DM), Betweenness Measure (BM), and Semantic Similarity Measure (SSM). Map/Reduce algorithm improves the efficiency of BET calculation especifically for BM and SSM calculation using the benefits of parallel processing. Finally, we use a Specific Interval(SI) to determine a set of Robust Expansion Terms (RET) and compare the result of our novel weighting approach with existing expansion approaches. We also show the effectiveness of our robust expansion in federated architecture.