A new weighted text filtering method

In order to describe the user´s profile exactly and provide what they need in information filtering, a new method to weight the user´s characterized demand and filter for him is put forward. In the paper, at first we define the terms´ weight of the user´s profile according to the order of the proposed sequence, then calculate the similarity between the user´s profile and the documents using the semantic dictionaries of Synonym Dictionary and HowNet, and at last submit the results to the user sorted in descending order of the similarity. The experiments show that it can improve the precision of information filtering.