中文网络评论的IT产品特征挖掘及情感倾向分析  

Mining Information Technology Product Features and Sentiment Orientation from Chinese Customer Reviews

作　　者：吴丽华[1] 冯建平[1] 曹均阔[1] 

机构地区：[1]海南师范大学信息科学技术学院,海口571158

出　　处：《计算机与数字工程》2012年第11期52-54,131,共4页Computer & Digital Engineering

基　　金：2012年度海南省自然科学基金立项项目(编号:612121);2010年海口市重点科技项目(编号:2010071)资助

摘　　要：为探索中文客户评论中的IT产品特征及相关情感倾向的挖掘,帮助IT生产商和服务商提高改进产品和服务质量,提高竞争力。该文将采用情感分析技术,提出基于客户感知价值的产品特征挖掘算法,实现对于评论中IT产品特征及其情感倾向的语义分析、动态提取和综合信息挖掘;并根据用户的关注权重将产品特征和情感倾向进行排列。采用从互联网下载的真实IT产品评论语料中进行实验,初步验证了该方法的有效性。At mining product features and polarity of customer′s opinion from Chinese online customer reviews,in order to help manufactures and service providers improve their IT products and services,and gain competitive advantages.This paper proposed an approach based on immune-based optimization algorithm which unsupervised feature extraction algorithm and supervised sentiment analysis technology to realize the summery information of customer reviews.And then ranked the work features and opinions based on customer attention strength.Several actual Internet products reviews were tested by an experimental model.Final results demonstrate the validity of the method.

关 键 词：网络客户评论 IT产品特征 评论挖掘 情感倾向分析 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]