AlgorithmforChinesetextcategorizationbasedonclassfeaturevectorrepresentation
基于类别特征向量表示的中文文本分类算法*

This paper used the approach to Chinese text categorization without word segmentation, expressing text features with bi-gram model. Compared with the classification models with word segmentation, the approach avoided complicated computation of word segmentation. To increase the accuracy of the approach, proposed an algorithm based on the class feature vector representation. And analyzed theoretically and verified experimentally the efficiency of the algorithm.