Research On A Domain Ontology-Based Intelligent Web Retrieval Model

With the coming of information society, the information requirement of people is growing. Many search engines become the main way for people to get information from Internet. Search engines are based on information retrieval technology. However, the traditional search engine can not fulfill the semantic web and intelligent information retrieval requirements and this leads to new researches. Among these researches, the application of semantic web based on ontology has caught many eyes. Among the Chinese Internet search researches, the semantic web applications have not have deep and significative development. This paper brings forward new design to improve Chinese intelligent search engine so as to fulfill the task of semantic intelligent information retrieval. A model system, StockOntoSearch (SOS), is built based on the stock domain ontology. The main contents and results of this paper are:A professional domanial search engine is designed to give out the advantage of semantic web in search engine sufficiently so as to provide professional and better web information retrieval than large-scale current search engine.More clues leading to user requirements are explored by the characteristics of user search behavior. These clues are used to help the semantic reasoning engine's work to analyze user requirement better, so as to enhance the search precision.Intelligent man-machine interface is designed to be more easily accepted by using the hierarchical faceted categories of ontology metadata and to provide heuristic search navigation by question excitation service. These are to enhance the quality of semantic analysis, provide intelligent service based on ontology out of a new aspect.This paper does some experiment on the StockOntoSearch system. From the data of the experiment, this paper validates the correctness of our design and theory search, and shows the enhancement to search precision and recall.