Supporting personalized new energy vehicle purchase decision-making: customer reviews and product recommendation platforms

The maturity of Industry 4.0 technologies such as the Internet of Things and cloud computing has accelerated the development of various platforms. In new energy vehicle (NEV) recommendation platforms, customer reviews have been well recognized for their ability to provide value-added information to customers who are interested in purchasing NEVs. However, the countless NEV reviews on different recommendation platforms make it difficult for consumers to select their preferred NEV. The existing NEV recommendation platforms also do not automatically perform fine-grained sentiment analysis of the product attributes contained in reviews. Consequently, they cannot provide personalized purchase recommendations for consumers. To this end, this study aims to propose a product purchase decision support method based on sentiment analysis and multi-attribute decision-making to improve the accuracy of personalized NEV recommendation platforms. Sentiment analysis was conducted on the attribute reviews of NEVs on a product recommendation platform. Subsequently, the positive, negative, and neutral sentiment ratios obtained based on sentiment analysis were regarded as q-rung orthopair fuzzy numbers. The ratios were then recognized as cumulative prospect theory (CPT) inputs. The prospect values of each NEV under each attribute were calculated and further aggregated into a Muirhead
mean operator to finally obtain the product rankings. This method was used to portray the consumers’ decision-making process considering various situations and irrational psychological factors (e.g., risk preference attitude). The results show that our proposed approach can enhance the decision-making support capacity of product recommendation platforms by providing sentiment analysis and capturing customers’ preferences for product attributes. Additionally, it can recommend more suitable NEVs to meet personalized customer requirements.