PARALLEL METAHEURISTIC FOR SEMANTIC BASED QUERY EXPANSION

Abstract

Word mismatch problem between the user’s query and the retrieved search results is one of the biggest problems facing the information retrieval(IR) field. Due to this problem, several techniques have been proposed such as query expansion(QE), which improves the IR performance by giving a more suitable extended query for users in comparison to the original query. In this work, we are looking for the best combination of the words in the extended queries using the semantic-based query expansion approach “ConceptNet”. To find these combinations we applied a low-level parallel metaheuristic Iterated local search(ILS) and a high-level parallel metaheuristic Accelerated particle swarm optimization(APSO) with Local search.