Research on data-driven group consensus decision-making of green methanol vehicle evaluation based on BERTopic text mining

Abstract

In response to climate change and energy crises, a scenario has emerged globally where multiple technical routes coexist, including electric, hydrogen, and methanol vehicles. Green methanol vehicles (GMV), a new path for future travel, the reported media data is easy to influence decision-makers’ (DMs’) evaluation on them. For the issue of GMV evaluation, this study proposes a data-driven group consensus decision-making model based on text mining. Firstly, the topics of related methanol vehicle in news reports are extracted through BERTopic text mining model, which employs a pre-trained transformer-based language technology to determine criteria and weights. Then, to address the problem of inconsistent results obtained from different centrality calculation methods in social network, a method for determining the DMs’ weights based on multidimensional advantages of centrality theory and water-filling theory is proposed. Furthermore, to uncover the psychological black box of DMs, the K-means method and bounded confidence model are employed to design a dynamic large-scale group consensus mechanism, ensuring effective integration of decision-making information and consensus achievement. Finally, the proposed model is used to evaluate of GMV, coal-to-methanol, gasoline, and electric vehicles. Discussions on production-living-ecological benefits for GMV are used to confirm the practicality of the proposed model.