摘要
随着社交媒体的发展,不断增加的在线产品评论正在极大地影响电子商务市场,使得评论挖掘成为商业界与学术界共同的热点话题。针对中文产品评论的特点,本文提出一种基于领域本体的建模方法,通过建立评论挖掘模型来对产品评论的基本评价单元——"特征观点对"进行识别。该建模过程以设计科学研究方法论为指导。首先在模型设计阶段,构建面向产品评论的领域本体;然后在模型实施阶段,提出基于本体的特征观点对识别方法;最后在模型评价阶段,通过实验对评论挖掘结果进行评价。实验结果表明,本文提出的方法与其他基于统计的方法以及基于语义的方法相比,在性能上有明显提高,对克服口语化严重和语法不规范等问题具有良好的效果。此外,通过特征观点对的识别与统计,使产品评论这种非结构化文本转化为机器可读的、能理解的结构化表达并得到具有一定商业应用价值的信息。
With the development of social media, the increasing product online reviews are greatly influencing electronic market, making review mining hot topic in both business and academic circles. According to the characteristics of Chinese online review, this research proposes a novel ontology-based modeling method to build review mining model and to identify the basic appraisal expression in online reviews-＂ feature-opinion pair （FOP）＂. Guided by the design science research methodology, this paper establishes a domain ontology for product review firstly, then designs algorithms to identify feature-opinion pair based on the ontology, and conducts several experiments to evaluate the proposed review mining model. Experimental results indicate that the performance of the proposed approach in this paper is remarkably better than the two baseline methods, a statistic method and a semantic method. Furthermore, through identifying and analyzing feature-opinion pairs, the unstructured product review is converted into the structured and machine-sensible expression, and provides valuable information for business application.
出处
《系统工程》
CSSCI
CSCD
北大核心
2013年第1期68-77,共10页
Systems Engineering
基金
国家自然科学基金资助项目(70971099)
中央高校基本科研业务费专项(1200219198)
上海市科技发展基金软科学研究博士生学位论文资助项目(12692193000)
关键词
中文产品评论
特征观点对
领域本体
评论挖掘模型
Chinese Online Reviews
Feature-opinion Pair
Domain Ontology
Review Mining Model
