Modeling tourists preferences through online reviews: an UPL-PT-MULTIMOORA-EDAS group decision method for wellness scenic spots selection

Online reviews provide adequate information to aid scenic spots in measuring tourist preferences. Modelling tourist preferences from online reviews has become an increasingly relevant and a practical concern in tourism. However, the massive and ambiguous nature of online reviews increases the difficulty of decision-making, meanwhile, psychological behaviours of tourists have an extensive influence on the decision-making results. In light of these facts, this study proposes an UPL-PT-MULTIMOORA-EDAS method. On the one hand, it enables tourists fully utilize the valuable information of online reviews; on the other hand, integrating psychological preferences can depict the reality of decision-making. Specifically, after crawling the online reviews, latent dirichlet allocation (LDA) and self-organized learning (SOM) are used to mine key attributes and segment the tourists, respectively. Moreover, two novel methods have been proposed for calculating the weight of attributes, which are on the basis of the frequency and position, respectively. Subsequently, prospect theory (PT) is introduced into the decision-making framework to reflect the psychological behaviours of tourists in tourism selection. Finally, we illustrate our proposed method through an example of Chengdu-Chongqing economic circle wellness scenic spots selection, the robustness and effectiveness of the proposed method are substantiated through pertinent experimental investigations and comparative analyses.