主管:中国科学院
主办:中国优选法统筹法与经济数学研究会
   中国科学院科技战略咨询研究院

Chinese Journal of Management Science ›› 2019, Vol. 27 ›› Issue (7): 167-176.doi: 10.16381/j.cnki.issn1003-207x.2019.07.016

• Articles • Previous Articles     Next Articles

Identifying Regional Demand Preferences from Online Reviews

WANG An-ning1,2, ZHANG Qiang1,2, PENG Zhang-lin1,2, NI Xin1   

  1. 1. School of Management, Hefei University of Technology, Hefei 230009, China;
    2. Key Laboratory of Process Optimization and Intelligent Decision-making, Ministry of Education, Hefei 230009, China
  • Received:2017-08-27 Revised:2017-12-13 Online:2019-07-20 Published:2019-08-01

Abstract: Due to the regional difference in customer needs,customer preferences of different region are not the same. Therefore, the identification of the relationship between customer preferences and regional characteristics has become the basis for decision-marking of regional management strategies. A framework of regional preference identification based on online product reviews is presented in this paper. Firstly, the product features are extracted from online reviews, and the sentiment polarity of product features is determined according to the emotion dictionary. Then, the product satisfaction is measured based on the feature sentiment. Finally, the hypothesis test method is used to identify the regional differences on customer satisfaction and sentiment polarity of product features. In order to verify the validity of the framework, the automotive product review data from autohome com are utilized for case studies. The experimental results show that customer satisfaction and sentiment polarity of fuel consumption, space, appearance and interior are significantly affected by the regional factors. The relationship between customer satisfaction, sentiment polarity of product features and regional characteristics is established to identify customer preferences in different regions and provide theoretical basis for regional design and marketing strategies of enterprises.

Key words: online reviews, regional preferences, product feature, sentiment analysis

CLC Number: