摘要
目前,国内外对在线评论特征词的研究很少考虑到卖家发布的商品标题和描述信息,这使得数据挖掘过程盲目,挖掘结果准确率不高。采用聚类分析方法,把商品标题和描述考虑进来,搭建三层挖掘模型对在线评论进行研究和分析,提出定位L-K-中心点算法。实验结果证明,该方法能提高挖掘的准确率,减少挖掘时间。
At present, title and description of goods are rarely considered in the research of online reviews at home and abroad, this makes the mining process blindly and mining results are not high accurate. In this article, the authors use the cluster analysis method, consider the title and description, set up a three - level mining model to analyze the online comments, at the same time, a location - clustering - algorithm is proposed. Experimental results show that the method improves the accuracy of mining and reduces the mining time.
出处
《现代图书情报技术》
CSSCI
北大核心
2011年第5期49-54,共6页
New Technology of Library and Information Service
基金
国家自然科学基金项目"面向协同的制造企业知识建模与集成理论研究"(项目编号:70871034)的研究成果之一
关键词
聚类
特征词
定位
K-中心点算法
Cluster analysis Signature words Location K - center - algorithm
作者简介
E—mail：252884250@qq．com
