A Possible Degree-Based D–S Evidence Theory Method for Ranking New Energy Vehicles Based on Online Customer Reviews and Probabilistic Linguistic Term Sets

As people’s environment awareness increases and the “double carbon” policy is implemented, the new energy vehicle (NEV) becomes a popular form of transformation and more and more car manufacturers start to produce NEVs. Thus, how to choose an appropriate type of NEVs from many brands is an interesting topic for customers, which can be regarded as a multiple-attribute decision-making (MADM) problem because customers often concern several different factors such as the price, endurance mileage, appearance and so on. This paper proposes a possible degree-based D–S evidence theory method for helping customers select a proper type of NEVs in the probabilistic linguistic environment. In order to derive decision information reflecting customer demands, online customer reviews (OCRs) are crawled from multiple websites and converted into five-granularity probabilistic linguistic term sets (PLTSs). Afterwards, by maximizing deviation and minimizing the information uncertainty, a bi-objective programming model is built to determine attribute weights. Furthermore, a possible degree-based D–S evidence theory method in the PLTS environment is proposed to rank alternatives in each website. For fusing these ranking results, a 0–1 programming model is set up by maximizing the consensus between the comprehensive ranking and individual ones in each website. At length, a case study of selecting a type of NEVs is provided to show the application and validity of the proposed method.