Abstract:Feature weighting algorithm has a great effect on the accuracy of text categorization.The classical Term Frequency and Inverse Documentation Frequency algorithm (TFIDF) ignores the semantic relationsh...Feature weighting algorithm has a great effect on the accuracy of text categorization.The classical Term Frequency and Inverse Documentation Frequency algorithm (TFIDF) ignores the semantic relationships between terms in the document set, thus to influence the accuracy of term weight calculation.To calculate the weight of words more correctly, the paper introduces the semantic association between words and proposed a new improved algorithm (S-TFIDFIGE) combined with semantic, information entropy and information gain.Experimental results show that the proposed method has better classification results than the traditional TFIDF and other improved algorithms.Read More