Query Semantic Expansion Model Based On Ontology And Mutual Information

As the developing of Internet Technology, human-being has stepped into the Information Apes. The freedom of delievering information and simplification of providing information accelerate the WWW's flourishing. The more information there are on Internet, the more difficult to find the proper information in such enviornment which be filled data in the format of non-structural and mul-structural.So it becomes to be a research hot field to find out the way to retrieve the related information high efficiently. Publishers always use different keywords or terms to describe the same knowledge point, even more, users always just type in one word to search information they need. Such phenomenon makes things worse.Undering such background, it to, be our research's purpose to make users get much more useful information they need. We proposed the Query Semantic Expansion Model Based on Ontology and Mutual Information (QSE_OMI) to achieve the purpose. Our researches as follows:1,Discuss how to combine the Mutual Information (MI) and Query Expansion (QE). MI is used to express the relation between two words; meanwhile QE is to expand those keywords which were inputed by users. In this paper, we build a QE model to combile those two theories after analysing and researching them;2,In this paper, we also describe the advantages of Ontology and MI. We also put them together because we found the advantages of their own can make up the defects of the other side. The another advantage of putting them together is they can bring the semantic into the QE and reduce the redundancy of expansion;3,We build the experiment system QSE_OMI upon those foundations we just mentioned above. After surveying the result of the system, it can improve the ratio of precision and recall of the information retrieve.The QSE_OMI we put forward can support the user profile model; it also can understand the semantic of keywords inputed. It has some forcement effection to the QE field.