Text categorization which assigns natural language texts to one or more predefined categories based on their content is an important component in many information organization and management tasks. Different automatic learning algorithms for text categorization have different classification accuracy. SVM classification model is common powerful for text categorization task. It is based on probability and is of religious theoretic basis. In this paper the SVM categorization model is analyzed and an algorithm to perform text categorization using incremental model is presented. Compared with the Bayes learning method and the K-nearest neighbor method experimental results verify the effectiveness of the proposed algorithm. Experiments show that the incremental model dramatically reduces the training time and is a better classification algorithm.