摘要
提出一种基于复杂网络的特征选择方法,通过构建文本加权复杂网络来表示词语间的语义关系及结构信息,综合考虑节点加权度、加权聚集系数、节点介数计算节点特性,利用节点综合特性提取反映文本主题的关键词作为文本的特征词。给出基于复杂网络的中文文本特征选择算法,并对其进行实验验证。结果表明,该特征选择方法较传统方法在文本分类性能上有所提高。
This paper proposes a feature selection method based on complex network. The weighted text is built to represent the semantic relations between words and text structure. The weighted degree, coefficient and betweenness are considered in the characteristics calculation of network nodes, the k complex network of weighted clustering ey words which can reflect the theme of the text are selected by the synthetic characteristics of network nodes. A Chinese text feature selection algorithm based on complex network is proposed and verified. The results of experiments show that the method proposed in this paper can get a better effect on the performance of text classification.
出处
《现代图书情报技术》
CSSCI
北大核心
2012年第9期23-28,共6页
New Technology of Library and Information Service
关键词
复杂网络
语义相关关系
节点综合特性
特征选择
Complex network Semantic relevance relation Synthetic characteristics of nodes Feature selection