Research Of Dynamic Knowledge Based Information Retrieval System

Information Retrieval (IR) is concerned with locating documents that are relevant for a user's query from a large collection of documents. There are two kinds of problems related to glossary in IR: one is "original expression"; the other is "difference of expression", which impact largely system's retrieval efficiency. Many approaches have been proposed to solve these problems, and each one has its advantages and disadvantages. On the basis of these approaches, we discuss and implement DKIRS (Dynamic Knowledge base Information retrieval System), which addresses these problems by carrying out the query expansion through dynamic knowledge. We adopt local context analysis to extract characteristic words from the documents retrieved by a user query, then apply subsumption approach and resemble approach in discovering terms relationships. The dynamic knowledge is composed of these extracted terms and term relationships, as well as user's feedback information. DORS expands user's query according to the dynamic knowledge. Experimental results show that the construction of the dynamic knowledge is reasonable and the retrieval efficiency is improved.