An AI-Driven Behavioral Analysis and Decision Support System for Smart Tourism Mini-Programs: Integrating UTAUT2 with Machine Learning

This study uses the UTAUT2 framework and machine learning to investigate the factors influencing Chinese tourists’ acceptance of smart tourism mini-programs. Social influence, enabling environments, hedonic motivation, performance expectancy, effort expectancy, and habit were found to be significant determinants of tourists’ behavioral intentions based on data from 504 respondents, according to structural equation modeling. Habit (β=0.539) and hedonic motivation (β=0.571) are regarded as the best predictors. These results were validated by additional testing with the Random Forest and XGBoost algorithms, which produced prediction accuracies of 88.5% and 89.2%, respectively. The results provide valuable insights and evidence supporting the adoption strategy for developers and tourism policymakers.