摘要
针对电力档案自动分类中应用效果不佳的问题,提出基于多特征选择的电力档案自动分类方法。首先,对电力档案文本内容进行提取、分词、去停词处理,并利用向量空间模型表示电力档案本文;其次,利用多特征选择技术提取文档频率、卡方检验、归一化差异、基尼指数及信息增益多项特征;最后,根据特征确定电力档案文档与类别的相似度,通过与分类阈值对比确定电力档案类别。实验结果表明,设计方法的档案错误分类数量较少,优于传统方法,在电力档案自动分类方面拥有广阔的应用前景。
A multi feature selection based automatic classification method for power archives is proposed to address the issue of poor application performance in automatic classification of power archives.First,the text content of power archives is extracted,word segmentation,stop word removal,and vector space model is used to represent the power archives text.Secondly,multiple feature selection techniques are used to extract multiple features such as document frequency,chi square test,normalized difference,Gini index,and information gain.Finally,the similarity between power archive documents and categories is determined based on their characteristics,and the power archive categories are determined by comparing them with classification thresholds.The experimental results show that the design method has a smaller number of file misclassification errors,which is superior to traditional methods and has broad application prospects in automatic classification of power files.
作者
马宁
李瑞环
MA Ning;LI Ruihuan(Shengzhou Power Supply Company of State Grid Zhejiang Electric Power Co.,Ltd.,Shengzhou Zhejiang 312400,China)
出处
《信息与电脑》
2023年第10期19-21,共3页
Information & Computer
关键词
多特征选择
电力档案
自动分类
multi-feature selection
power file
automatic classification
