摘要
将基于等价关系的模糊聚类技术应用于中文文本分类,提出了基于模糊聚类的中文文本分类算法ATCFC。该算法利用基于二级字索引的正向最大匹配算法对文本分词,建立模糊特征向量空间模型,使用贴近度法刻划文本间的相似度。利用算法ATCFC对文本集合进行动态聚类实验,实验结果表明算法ATCFC对于中文文本分类是可行、有效的。
This paper studies Chinese text categorization with the technique of fuzzy clustering based on equivalence relation and proposes an algorithm(ATCFC) for Chinese text categorization based on fuzzy clustering, This algorithm uses forward maximum match algorithm based on two-level word-index to segment Chinese text,creates fuzzy feature vector space model and describes similarity degree among texts using the method of close degree.Algorithm ATCFC is used to conduct a dynamic clustering experiment on a text set and the experimental results demonstrate that algorithm ATCFC is feasible and effective for Chinese text categorization.
出处
《计算机工程与应用》
CSCD
北大核心
2006年第8期170-172,177,共4页
Computer Engineering and Applications
基金
江苏省重点实验室开放基金资助项目(编号:KJS03064)
关键词
模糊聚类
文本分类
贴近度
模糊等价矩阵
fuzzy clustering, text categorization, close degree, fuzzy equivalence matrix