Efficient inverse query expansion in information retrieval using knowledge reduction

Query expansion has been used in information retrieval (IR) to improve the performance of retrieval results. Even with the advantages of query expansion, most of the methods rely on costly query processing such as relevancy feedback. We suggest efficient method of query expansion for removing query processing when a user requests search. The method called "reverse query expansion" expands document vectors instead of a query separate from retrieval process. This method produces same effect as a query expansion while executing comparable speed with the vector IR model.