An algorithm of text classification is given that imitates human's in this paper. On one hand, the algorithmenhances weight of theme when feature vector is processed, because of the assumption that the title of a document canproject its content. On the other hand,a weight parameter o vector is designed to simulate human's skimming andskipping behavior for calculating method of a document cluster center, and a weight of the feature that there are morepositive examples than negative ones is enhanced . The experiment shows that the algorithm greatly improves the per-formance of a text classification system.
