A New Method for Applicant of Explicit Semantic Analysis and Word Sense Disambiguation in Concept-based Information Retrieval

previous Information retrieval (IR) systems based on keywords to retrieve and index documents. They may return inaccurate results when different keywords are employed to illustrate the same concept in the documents and in the queries presented by Users. In Concept-based retrieval methods have tried to tackle these troubles by using concept-based comparison between documents and queries. Therefore, accurate concept extraction of documents and queries improves performance of IR systems. In this research, we introduce a new concept-based query semantic analysis approach based on Wikipedia-based Explicit Semantic Analysis. We propose that first specify the given context of query words by using Wikipedia-based concept network named wikinet to query words sense disambiguation. Then rely on given context create related concepts of query that they will be compare to concepts of documents. Because the main aim of this paper is to provide a correct interpretation and semantic relatedness analysis of query words, we use of correlation of computed relatedness scores with human judgments. Evaluation shows that the proposed method provides improvements compared to the existing semantic analysis methods. Keywordsconcept-based information retrieval, query, query sense disambiguation,