《词分布均衡度评价法》实验综述

Experimental Survey of Word Distribution Equilibrium Degree Evaluation

支持向量机文本分类因其分类精度高而得到广泛应用,本文提出了基于词分布均衡度支持向量机文本分类算法的实验要求、实验条件、实验步骤及实验结果分析。实验结果表明,在数据挖掘的文本分类中词分布均衡度评价特征词选取法优于优于标准文档频数法等方法。

Text categorization with support vector machine is widely used because of its high classi- fication accuracy. The experimental requirements, conditions, procedures and results analysis based on the algorithm of text categorization with support vectormachine of word distribution equilibrium degree are proposed. The experimental results show that for feature word selection in data mining, the evaluation method of word distribution equilibrium degree is better than standard document frequency method and other methods.