Study On The Online Tourism Product Selection Based On Granular BWM Group Decision Method

In recent years,with the improvement of Chinese residents’ income level,people’s pursuit of quality of life is not limited to the material level,but began to change to the spiritual and cultural level,and the demand for tertiary industries such as tourism service industry is increasing rapidly.Simultaneously,due to the popularity of mobile payment,more and more people tend to purchase online tourism products.However,the advent of the era of Internet of things also makes the data grow explosively.When consumers choose online tourism products,they often spend a lot of time and energy to browse the corresponding online reviews.To a certain extent,it increases the decisionmaking cost and reduces the decision-making efficiency.Therefore,this thesis introduces a knowledge and data collaborative driven group decision-making method,that is,the BWM(Best-Worst Method)model based on granular computing,to explore the selection of online tourism products in group decision-making problems.Through the integration of uncertain decision theory and granular computing theory,this study constructs a trinity research framework of "Construction of online tourism products criteria evaluation system—Calculation of online tourism products criteria weights — Multi-criteria decision-making method of online tourism products selection",which helps consumers provide scientific decision support when choosing online tourism products.Due to the diversity types of online reviews,this study establishes the corresponding decision-making models to deal with different types of information.Firstly,for the numerical online reviews,interval-based information granules are constructed,and the adaptive particle swarm optimization(APSO)algorithm is applied to determine the optimal distribution of information granules,so as to obtain the decision result in the level of highest consistency.Secondly,for the linguistic online reviews,this thesis considers interval-based and interval type-2 fuzzy sets-based information granules to granular the linguistic terms.Finally,the principle of justifiable information granularity is introduced,combined with the specificity and coverage of information granularity.A granular neural network is established to gather group decision information and promote the process of group consensus reaching.Owing to the randomness of online reviews,this study uses Monte Carlo stochastic simulation method to randomize the decision weight of each review to increase the robustness of this model.In the group consistency optimization and group consensus reaching process,the iterative convergence algorithms are constructed to adjust the evaluation information.This study collected online tourism products reviews from websites such as Tripadvisor.cn,Booking.com and Dianping.com.Then,discussed and analyzed the real cases of tourist spots selection and tourist hotels selection,and verified the effectiveness and feasibility of the proposed model.The innovations of this study are as follows: firstly,this thesis uses Python served as the data collection tool to obtain online tourism products reviews,establishes a datadriven tourism products criteria system,combines granular computing theory with multi-criteria decision-making method,and proposes a BWM model based on the optimal allocation of information granularity to calculate the criteria weights.Secondly,in the evaluation process of online tourism products,the idea of machine learning is input to construct a granular neural network to optimize the consistency and consensus in the decision-making process.Thirdly,on the basis of the the above theory and method innovation,the robustness and practicability of this model are explored through making the sequence of tourist spots and the selection of tourist hotels as the actual application scenarios.It provides a new research perspective and useful reference for data-driven online tourism decision-making analysis.