Cross-subject page ranking based on text categorization

With the development of Internet, there are enormous web pages in the Internet. So the good page ranking algorithm is critical for users to gain positive results. The traditional ranking method is suitable for general search engine, but not for the focused search engine and the search engine based on categorization. With state of the art in text categorization, so many cross-subjects appear, and the cross-subject web pages also exist in search engine. When we retrieve the cross-subject web page, they pages which satisfy the users´ demands will appear at last of result lists, because their score is lower than subject web pages. This paper mainly discusses the problem of cross-subject page ranking problem. After analysing the traditional page ranking algorithm, we proposed a new method named categorization-based ranking algorithm which can enhance the score of cross subject web pages. This method optimizes the order of the result list, and improves the quality of search engine.