Web celebrity shop assessment and improvement based on online review with probabilistic linguistic term sets by using sentiment analysis and fuzzy cognitive map

Abstract

As a representative of the new economy, the web celebrity economy has achieved significant development in China with the rapid development of information technology and the Internet. In this environment, web celebrity shops encounter fierce business competition of peer competitors. Online reviews which imply the consumers’ attitudes and sentiments give the web celebrity shops good feedback to improve their competitiveness. Thus, taking milk tea as an example, this paper deeply investigates the assessment of web celebrity shops by mining online review. At the same time, we also discuss the competitive analysis and propose the corresponding improvement advices. In order to obtain the satisfaction assessments of web celebrity shops, on the one hand, we analyze topic extraction with latent dirichlet allocation (LDA) and determine the attributes that customers care about. On the other hand, we utilize long short-term memory (LSTM) and probabilistic linguistic term sets (PLTSs) to more precisely portray customers’ sentiment towards different attributes. By using fuzzy cognitive map (FCM) and the association rule, we further investigate the interrelationship among the attributes and construct the relationship graph between attributes for web celebrity shops. With the above results, we aggregate the decision information by designing improved extended Bonferroni mean (EBM) and obtain comprehensive evaluations. General speaking, this paper successfully transforms the unstructured data of online reviews into quantitative information and obtain satisfaction evaluations. With the aid of PLTSs and FCM, we further investigate the competitive analysis and propose improvement advices for each shop, which systematically provides us with a data-driven decision-making analysis model.