Evaluating Quality of Chinese Product Reviews Based on Fuzzy Logic.

Wei Wei,Yang Xiang,Qian Chen,Xin Guo
DOI: https://doi.org/10.1007/978-3-642-24273-1_45
2011-01-01
Abstract:The prevalence of web2.0 makes e-Comtnerce an increasingly popular trend. Consumers can post their product reviews about product after buying products on the web. These product reviews can help online user make sensible decisions and enable business enterprises to improve their business strategies. To cope with the information overload problem, opinion mining is needed to extract useful expression and summarize important opinion for users. But the quality of product reviews at websites varies greatly. In this paper, we propose a method based fuzzy logic to evaluate the quality of Chinese product reviews. We define three fuzzy sets to represent the different types of product reviews. We determine the quality of product reviews under the proximity membership principle based the three fuzzy sets. Experiments based on an expert-composed product reviews corpus show that our method can achieve promising performance.
What problem does this paper attempt to address?