摘要： 分析贝叶斯文本分类算法的不足，提出相应的改进算法。放宽朴素贝叶斯文本分类模型中的属性独立性假设，采用一种改进的基于贝叶斯定理的文本分类模型“树桩网络”，改进朴素贝叶斯文本分类模型。实验证明，改进后的文本分类模型适合于文本分类的需要，改善了原有分类器的性能。

关键词: 文本分类, 朴素贝叶斯, 属性独立性假设, 树桩网络

Abstract: This paper analyzes the shortcomings of Bayes and puts forward a better method to improve it. It releases attribute independence assumption of Naive Bayes text classifier. An improved text classification model based on Bayes theorem called stump network is presented to amend the Naive Bayes text classifier. Experiment shows that the revised text categorization model meets the need of text categorization, and improves the performance of former one.

Key words: text classification, Naive Bayes, attribute independence assumption, stump network

中图分类号: 

• TP311