网购评论情感数据的k近邻和SVM处理方法研究  

Research on online-shopping sentiment based on k-nearest neighbor and SVM algorithm

作　　者：焦畅[1] 佘玉梅[1] 丁冬冬[1] 庄丽[1] 

机构地区：[1]云南民族大学数学与计算机科学学院,云南昆明650031

出　　处：《云南民族大学学报（自然科学版）》2013年第3期209-212,共4页Journal of Yunnan Minzu University:Natural Sciences Edition

基　　金：国家自然科学基金(08XMZ002)

摘　　要：针对网购评论,抽取评论语组成基本语料,构建客户网购评论情感词汇本体,对热点评论应用k-近邻和SVM 2种算法来分析评论文本热点事件,实验证实SVM算法较k-近邻算法在评论文本热点发现上具有较高的性能,为网购评语热点研究提供了实例参考.Based on the basic corpus composed of comment extracts from online - shopping, the thesis builds a voabulary database of customers＇ attitudes towards the online - shopping comments. The comments on the hot events are analyzed by applying k - nearest neighbor and SVM algorithm. The result shows that SVM algorithm is better than k -nearest in this aspect, which provides reference for studying on the online -shopping comments.

关 键 词：网购评论 K近邻 SVM算法 意见挖掘 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]