Research On Text Information Retrieval Based On Ontology

From lacking of information in the beginning to the current information overload the World Wide Web has evolved rapidly, the amount of information on the internet is expanding so quickly that, there has never had such a flood of information in history, now, the emergent task is how to obtain useful information efficiently and exactly. The emergence of information retrieval technology will profoundly affect the way people get information.In the practical application, the most search results of current retrieval system are often unsatisfactory. The reason is that these search engine are based on keyword matching in which, the meaning and the concept of the keyword are not at the same level, so the results of the query matching only in the literal meaning. The current retrieval system based on keywords expansion which concerned was the keyword's synonyms. But, the concept and relevant knowledge of the word are needed by people.Existing search technology has made brilliant achievements, this is not only because of its successful operations, largely but also because of the classical algorithm behind. The paper draws the conclusion that ontology technology can be used to enhance the search satisfaction level of the information retrieval. By experiment this paper shows that the recall and precision rate of retrieval can be improved largely by this method.The paper has done following several studies:①Studied the ontology construction methods and its application areas.②And studied three typical retrieval system models and its search algorithms, then summarized the advantages and disadvantages of each model, and their application areas.③After studied the classic information retrieval algorithms deeply ,we upgraded it to new algorithm by using ontology base, the search algorithm of classic retrieval system was improved, and in the retrieval process joined the background knowledge expansion of keywords.④The paper studied user's query conditions, the system classified the user's query conditions through the ontology knowledge, further defined the classification and limited its range of user query.