基于数据挖掘中文书目自动分类算法

Automatic classification algorithm of Chinese bibliography based on Data Mining

摘要:

提出一种改进的数据挖掘算法。首先采用ICTCLAS系统进行文本预处理,以词频特征构建词条向量；然后融合词频特征和词频-逆向文件频率特征,构建训练样本集的特征矩阵；接着对该矩阵进行奇异值分解变换,得到语义空间,用于对文本特征向量进行语义空间变换,得到语义向量；最后构建联合支持向量机分类器,实现中文书目所对应的语义向量的自动分类。最后做了大量的仿真实验,实验结果表明,本文方法的分类准确率高于现有方法。

Abstract:

an improved algorithm for data mining is proposed. The first use of ICTCLAS system for text preprocessing, construct the term vector in frequency characteristics; then the fusion frequency characteristics and frequency - inverse document frequency features, construct the characteristic matrix of the training sample set; then the matrix singular value decomposition, get the semantic space for semantic space transform of text feature vector, semantic vector; the construction of combined support vector machine classifier, automatic classification of semantic vector corresponding to the Chinese bibliography. At last, a lot of simulation experiments have been done, and the experimental results show that the classification accuracy of this method is higher than that of the existing methods.