Identification and evaluation of competitive products based on online user-generated content
Ya'nan Dong,Yupeng Li,Jin Cao,Na Zhang,Kaixin Sha
DOI: https://doi.org/10.1016/j.eswa.2023.120168
IF: 8.5
2023-04-22
Expert Systems with Applications
Abstract:In online shopping, competitive product identification and evaluation (CPIE) are becoming indispensable for customers before making purchase decisions. Online user-generated content (UGC), as a common way of expressing customers' opinions or preferences toward products, is a key proxy for CPIE. Some studies have performed CPIE by the descriptive information in online reviews, while still lacking the exploration of comparative information, which is essential for potential customers to compare alternatives and usually embedded in online Questions & Answers (Q&As). In this study, a framework for CPIE based on multi-type information in online UGC is proposed. First, the competitive products toward a target product are identified from the comparative Q&As according to the co-occurrence of product names. Second, the descriptive information and comparative information which derived from online reviews and comparative Q&As, respectively, are integrated by constructing a comparative network. The nodes and edges in the comparative network represent the products and competitive relationships, respectively. The holistic evaluation of competitive products is accomplished through the PageRank algorithm. Finally, the empirical analysis is illustrated to verify the feasibility and effectiveness of the proposed method, and a system prototype of the CPIE is designed and developed to facilitate the purchase decisions for customers.
computer science, artificial intelligence,engineering, electrical & electronic,operations research & management science