Tourism attraction selection driven by online tourist reviews: A novel multi-attribute decision making method based on the evidence theory and probabilistic linguistic term sets

Abstract

In today’s internet age, potential tourists often browse online tourist reviews (OTRs) before determining travelling destinations. However, the information in massive OTRs is usually fuzzy and uncertain. The probabilistic linguistic term set (PLTS) is a helpful tool to describe the ambiguous and uncertain information. This study utilizes the PLTS to depict information in OTRs and proposes a novel tourist attraction selection method with OTRs. First, a new score function of PLTSs is introduced by combining risk attitudes of decision makers (DMs) and the hesitancy of the PLTS. Subsequently, the attributes evaluating tourist attractions are determined by extracting the top 50 high frequency words from OTRs. A new sentiment analysis technique is developed for transforming OTRs into PLTSs with the five-granularity linguistic term set. According to the decision information and the number of times attributes commented, a bi-objective programming model is built to derive attribute weights. Finally, fusing alternative perceived utility values into the D-S evidence theory, a new decision method is developed to rank alternative tourist attractions. At length, a case study of selecting tourist attractions is provided to illustrate the applications of the proposed method. Furthermore, the comparative analyses are conducted to show its effectiveness and superiority.