Design And Implementation Of Ontology-based Classification Retrieval System

With the rapid development of the information era, the amount of information on the internet is becoming more and more information, though, but people are online to find the information is becoming more and more difficult. Though some search engines have provided tools to effectively search information from enormous web pages, but, looked from the present domestic and foreign research condition, the existing search engine uses the retrieval system is very difficult satisfyingly, the reason mainly lies in: The retrieval system question classification causes the final answer and the question subject deviation not precisely; The existing answer extraction correlation technology mostly based on the statistical method, has neglected the sentence semantics, like this can cause the inquiry result existence massive useless and the redundancy information, therefore has affected the answer rate of accuracy. So if enhance the existing retrieval system the precision, we must solve the problem which above mentioned.This paper proposed a information retrieval system to make the results more abundant and more accurate by utilizing ontology knowledge. Ontology is a kind of model that is used to describe the concepts and the relations of them, and ontology can give between the glossary and the glossary from in the different level formalized pattern the reciprocity is clear about the definition. Field ontology includes abundance knowledge and semantic relations of this field. Through applying these resources to information retrieval system by a certain way. It is impossible to solve the problem of semantic understand insufficiency in a way.Based on this, this paper puts forward a method of classification retrieval system based on ontology, using lots of ontology semantic relations, from the semantic level reveals the relationship between document itself and document, classify users'questions so as to enhance information retrieval precision rate and recall rate. The system of functional modules include three parts: the use of experimental fields, storage and maintenance; classify the problem sets preprocessing; pretreatment of user problems, namely, classification, expand, and will eventually inquires retrieval results returned. This system will run the main can be divided into a few steps: for problem sets, and through the analysis of classification procedures and ontology concept map; resolve the ontology files and calculate the various concepts weight, according to different concept tree structure stored in a database; user input inquiries and submit; analysis the user's questions and provide the idea about the concept of common sense; by calculating the user questions and candidate question, users according to results conform to decide whether to once again.