Query Expansion Based on Mongolian Semantics

Accurately and effectively measuring the similarity between terms is one way to enhance the performance of information retrieval systems. At present, the extended data source which is used for Mongolian semantic query expansion is still not perfect. A new query expansion method based on Mongolian semantics is proposed for expressing the user´s intention more precisely. This query expansion method includes two steps. The first step is to build an associated-dictionary by associated term rearrangement algorithm. The second step is to expand the original query by the associated-dictionary. The core of the method adopted in this paper is associated term rearrangement algorithm and re-weight expansion terms algorithm. Experiments with Lemur tools show that the retrieval results have been greatly improved by this method.