网络舆情安全应用中主题分类方法的研究与实现  

Implementation of Subject Classification on Network Public Opinion in Public Safety

作　　者：马海兵[1] 毕久阳 邱君瑞[2] 

机构地区：[1]武警政治学院政治工作系,上海200435 [2]第二军医大学图书馆,上海200433

出　　处：《现代情报》2012年第4期8-13,共6页Journal of Modern Information

基　　金：2010年度国家社会科学基金项目"群体性突发事件网上监测预警平台及其关键技术研究"(编号:10CTQ024)成果之一

摘　　要：针对网络舆情安全应用中主题分类问题展开研究。主题分类需要解决5个问题。建立了网络舆情安全应用的分类体系;介绍了文档特征提取和选择方法;分析了向量空间模型;实现了KNN和SVM两种分类算法。验证了方法的实用性。The problem of subject classification on network public opinion in public safety was addressed in this paper. Subject dassificafion need to solve five issues. Firstly, a classification system was established on network public opinion in public safety. Secondly, methods of document feature extraction and selection were introduced. Thirdly, the vector space model was analyzed. Then, Two kinds of classification algorithms, KNN and SVM were implemented. Finally, the result of experiments were proposed.

关 键 词：网络舆情安全 舆情分析 主题分类 

分 类 号：G254.1[文化科学—图书馆学]