Tourist Attraction Recommendation Model Based on RFPAP-NNPAP Algorithm

Abstract

Driven by globalization and digitization, the tourism industry is facing new challenges and opportunities brought about by big data and artificial intelligence. The recommendation of tourist attractions, as an important part of the industry, has a direct influence on the tourist experience. However, with the diversification and personalization of tourism demand, traditional recommendation methods have shown shortcomings: weak processing ability for complex nonlinear data, affecting recommendation accuracy and personalization, and insufficient efficiency and stability when processing large-scale data. Faced with this challenge, this study proposed a hybrid tourist attraction recommendation model with random forest, artificial neural network, and frequent pattern growth. This model utilized the powerful classification and regression capabilities of random forests, as well as the complex nonlinear mapping ability of artificial neural networks, to predict tourist attraction preferences. And on this basis, the frequent pattern growth algorithm was introduced to mine the associated attractions of tourist preferences, thereby achieving accurate recommendation of tourist attractions. In experimental verification, the proposed model demonstrated superior performance. It not only surpassed traditional tourist attraction recommendation methods in accuracy and personalization, but also exhibited efficient and stable characteristics when processing large-scale data. After about 16 iterations, the MAPE value of the mixed model decreased to 0.44%. After about 39 iterations, the MAPE value of the mixed model decreased to 0.40%. The average accuracy, recall rate and F-value of the proposed model are 92.26%, 82.11% and 84.43%, respectively, which are superior to the comparison algorithm. Its error correction accuracy fluctuates around 90%. This study provides a new solution to the problem of personalized recommendation of tourist attractions, providing theoretical guidance for the tourism applications of random forests and artificial neural networks, and improving the tourist experience, promoting the development of the tourism industry.