您的位置:首页 > 论文页面
基于产品特征聚类的产品评论信息挖掘研究
发表时间:2014-02-28 浏览量:2550 下载量:804
全部作者: | 苏红旗,聂金慧 |
作者单位: | 中国矿业大学(北京)机电与信息工程学院 |
摘 要: | 产品评论中的产品特征及情感倾向是一种反映产品价值的重要信息,针对现有的中文产品评论信息挖掘存在的不足,提出一种基于词汇共现性的产品特征聚类技术与细粒度情感分析技术。在产品评论中,同类产品属性会有多种多样的表述方式,研究把产品评论中的产品特征进行归类,且与以往基于句子的整体情感分析不同,提出针对产品特征的更细粒度情感分析技术,并且对没有相应情感倾向的属性词做出合理处理。 |
关 键 词: | 计算机应用;产品特征归类;相邻共现;细粒度情感分析;非极性观点词 |
Title: | Information mining of product comment based on product feature cluster |
Author: | SU Hongqi, NIE Jinhui |
Organization: | School of Mechanical Electronic and Information Engineering, China University of Mining and Technology |
Abstract: | Product features and emotional tendencies in product reviews reflect the value of things. According to the deficiency of Chinese information mining, we present a product feature cluster technology and fine-grained emotion analysis technology based on the vocabulary co-occurrence. In product comments, one feature may have various of means of expression, this paper classifies the features in product comments, presents a more fine-grained emotion analysis technology different from former emotion analysis that based on sentences. In addition, reasonable processing method is presented for the features without corresponding emotion tendency. |
Key words: | computer application; cluster of product features; co-occurrence; fine-grained emotion analysis; non-polar viewpoint |
发表期数: | 2014年2月第4期 |
引用格式: | 苏红旗,聂金慧. 基于产品特征聚类的产品评论信息挖掘研究[J]. 中国科技论文在线精品论文,2014,7(4):332-338. |

请您登录
暂无评论