IMPROVING THE VECTOR SPACE RETRIEVAL MODEL BY USING TERMS'S MODIFIERS AND CONSTRUCTING FUZZY SYNONYM THESAURUS

摘要

An information retrieval (IR) system can facilitate users to quickly and efficiently find out the documents that are relevant to the users' requirements. However, most current keyword-based IR methods often generate large trashes and miss important information since separated words/terms are involved in the retrieval process. Treating words/terras in the given query separately will lead to a number of irrelevant documents retrieved. To deal with this problem, we accordingly propose a new method in which the relationships among words are fully considered. For expanding the user's query, the fuzzy synonym thesaurus is built. Through the proposed method, we can retrieve relevant documents in a relatively narrow search space and meanwhile widen the coverage of the retrieval to the related documents that do not necessarily contain the same words as the query. As a result, the retrieving results obtained from the modified keyword-based IR method show some improvements in two metrics, precision and recall.