Information Discovery based on Multi-granularity Text Fusion

In this paper we introduce a new information discovery algorithm Multi-granularity Text Fusion (MGTF) on the Web. Granularity means the length of News relevant web documents, such as News web pages, Blog and Micro Blogs, which comes from web uses. The longer the text is, the higher of the granularity it has. Given a topic query on the Internet and the results of different granularity and time-stamped web documents which contain the query keywords, the task of MGTF is to orderly return those different granularity web documents discussed about the same topic. The process of multigranularity web documents analysis leads to heretofore unknown information and opinions that valuable potential, minority and contentious respectively, which integrates the time, content, reprint and link information. Experiments show that MGTF achieves the best overall performance with high effectiveness and robustness.