Research on Web Information Filtering Based on Incremental Feedback Algorithm

Whether the category template is good or not is directly related to the precision of the information filtering in the system. And the classification system may often be changed. This article presents an algorithm which modifies the category template according to the feedback documents. Experimental results show that this algorithm can make category templates much closer to the real ones as soon as possible and the algorithm is effective. Establishment of the initial category templates, expression of documents and the other techniques of the information filtering are researched in detail in the paper.