摘要
提出了一种评价对象特征抽取与聚类方法,其基本思想是:首先运用Apriori算法和剪枝方法从客户评论中抽取评论对象特征集;然后,以特征之间的基于Hownet的词语相似度和特征与观点共现的信息作为聚类依据,对特征进行聚类。采用通过从互联网获得的客户评论语料对该方法进行了实验,实验结果验证了该方法的有效性。
This paper proposes an approach to extract and cluster the features of evaluation object. The approach, at first, extracts features of evaluation object in user reviews by Apriori algorithm and pruning algorithms, and then, cluster features based on term similarity and co-occurrence between opinion words and features. With the Chinese reviews of mobile, digital camera, CD writer, cloth and book from the Internet, experimental results demonstrate the validity of the proposed method.
出处
《微型机与应用》
2014年第17期72-75,79,共5页
Microcomputer & Its Applications
关键词
关联规则
特征抽取
特征聚类
语义相关度
相邻共现
association rules
feature extraction
feature clustering
term similarity
co-occurrence
