基于CBC-LIKE算法的产品特征词聚类的研究  

Research on product feature words clustering based on CBC-LIKE algorithm

作　　者：江伟[1,2] 路松峰[3] 杨莉萍[4] JIANG Wei;LU Songfeng;YANG Liping(Wuhan University of Technology,Wuhan 430070,China;City College,Wuhan University of Science and Technology,Wuhan 430083,China;Huazhong University of Science and Technology,Wuhan 430074,China;Huazhong Agricultural University,Wuhan 430070,China)

机构地区：[1]武汉理工大学,湖北武汉430070 [2]武汉科技大学城市学院,湖北武汉430083 [3]华中科技大学,湖北武汉430074 [4]华中农业大学,湖北武汉430070

出　　处：《现代电子技术》2017年第14期81-84,共4页Modern Electronics Technique

基　　金：国家自然科学基金资助项目(61173050);中央高校基础研究经费资助项目(2662015QC040)

摘　　要：用户评论中存在产品特征表达多样性问题,在细粒度观点挖掘任务中需要对产品特征词聚类。首先,结合不同的语义相似度计算的特点,提出基于语义知识和上下文熵模型的语义相似度混合计算方法,计算抽取得到的特征词语义相似度;然后改进了传统CBC算法,提出适用于产品特征词聚类的CBC-LIKE方法实现聚类。最后在三个领域的真实评论语料上进行实验,对提出的语义相似度计算方法和聚类算法的性能进行了分析。实验结果表明,所提方法是有效的,与另外两种基线方法相比性能较优,取得了较好效果。Aiming at the problem of the various product feature expressions existing in user reviews,it′s necessary to cluster the product feature words in the task of fine-grained opinion mining.According to the calculation characteristics of differentsemantic similarities,a semantic similarity mixture calculation method based on semantic knowledge and context entropy modelis proposed to calculate the extract the semantic similarity of feature words.The traditional CBC algorithm is improved.A CBC-LIKE method suitable for product feature words clustering is proposed.The experiment is conducted for the real review corpus inthree domains.The performances of the proposed semantic similarity calculation method and clustering algorithm are analyzed.The experimental results show that the method is effective,its performance is better than that of other two benchmark methods,which has perfect effect.

关 键 词：产品特征 语义相似度 聚类算法 观点挖掘 

分 类 号：TN911-34[电子电信—通信与信息系统]