Research on Intelligent Analysis Model of Tourism Consumers’ Purchase Intention and Its Application Practice in E-commerce Platforms

Tourism consumption, as a special type of daily consumption, has significant features that distinguish it from other types of consumption. The purpose of this paper is to summarize and extract the feature engineering in tourism consumption, encode the feature engineering in tourism consumption behaviors using the Transformer model, and predict consumer purchase intention. After the performance evaluation of the Transformer-based consumer purchase intention analysis model constructed in this paper, it is applied to the e-commerce platform, taking the dynamic pricing strategy of the e-commerce platform as the entry point and analyzing the application effect of the model in the optimal inventory time path, the optimal price of the platform, and the expected revenue. The transformer-based consumer purchase intention analysis model in this paper is significantly better than other prediction models in AUC, accuracy, precision, recall, and F1. When applied to e-commerce platforms, this paper’s model helps to realize the inventory equilibrium state between inbound enterprises and e-commerce platforms, which improves the overall revenue of both parties. When time t=40 or inventory n=30, the optimal online store price of the platform after the model is lower than the price before the model is applied, and consumers can enjoy a better purchasing experience. t=40 or n=30, the total expected revenue of both parties after the model is higher than the total expected revenue before the model is applied. This paper shows that the Transformer-based consumer purchase intention analysis model can positively impact the dynamic pricing of e-commerce platforms.