摘要
本文设计了一个有效的基于贝叶斯分类器的中文期刊自动分类系统。首先,该系统以期刊的名称作为惟一的标引内容,并利用自动分词技术将期刊名称分成待分类的样本集;其次,通过对图书馆的样本数据进行训练建立的分类库,本文使用贝叶斯分类器实现中文期刊的自动分类。实验结果表明,该分类器对中文期刊的分类具有很好的高效性和准确性。
This paper presents an efficient automatic categorization system for Chinese journals based on Bayes classifier. First, the system uses the rifles of Chinese journals as the sole keyword indexes, and cuts them into several sample collections by word cut technic. Then, giving the sample words from library, a categorization database is created and used for automatic categorization of Chinese journals by Bayesian learning. The experimental results show that the Bayesian classifier can be successful in categorization for Chinese journals efficiently and correctly.
出处
《现代情报》
北大核心
2007年第4期146-147,150,共3页
Journal of Modern Information
关键词
中文期刊
分词
贝叶斯分类
Chinese journals
word cut
Bares classifier
