The Research Of Subject-oriented Search Engine Based On Lucene

Search engine collects and detects informations in the Internet by using certain strategies. Then search engine can understand, organize and handle the informations and provide information retrieval services. So search engine can help us to navigate informations in the Internet. Theme of search engines only cover specific topics related to regional Web. It can search deeper, search the cycle can be shorter. So that it can satisfied with users the performance requirements of accessing to information resources quickly and accurately. The study of theme search engine is in a very active stage at present. Much knowledge of machine learning field is applied to the design and implemention of theme search engine.A algorithm of spider searching strategy based on comprehensive values is brought out in this article. The algorithm combines the evaluation method based on immediate value with the evaluation method based on future value, and forecast the importance of the links using both of them. It enhances the efficiency of the theme search engine. A general framework is given for the design of the indexer and the detailed design of indexer, summary generator and abstract content shader have been completed. Finally, the Lucene original pages sorting algorithm is improved based on specific needs.