Modeling personalized cognition of customers in online shopping

Abstract

Abstract Online reviews play an important role in online shopping. Previous studies on Multi-Criteria Decision Making (MCDM) based on customer reviews focused too much on sentiment words in text reviews but ignored the personalized semantics of linguistic terms. In addition, only considering the qualitative information of products/services is not enough to simulate the purchase behaviors of customers given that the customers are also concerned with quantitative parameters. To bridge these research gaps, this study models the personalized cognition of customers on both quantitative and qualitative information, and proposes an MCDM framework for online shopping. Firstly, we determine the personalized semantics of linguistic terms through the emotional consistency between star ratings and text reviews. Afterwards, we investigate the “psychological intensity” based on Weber-Fechner s law to determine the utilities of quantitative parameters. A utility-based translation method is then developed to express both quantitative parameters and text reviews as probabilistic linguistic term sets. The unified information is further aggregated to represent the performance of products/services. The applicability of the proposed method is illustrated by a case study of television selection from Amazon.com. The results demonstrate that the personalized cognition has an influence on the judgments of products/services.