Explaining and Predicting Behavioral Intention Toward the Halal Phuket AI Tourism Platform: A UTAUT–Machine Learning Hybrid Study

Abstract

This study examined behavioral intention toward Halal Phuket, an AI-enabled tourism platform supporting Muslim travelers in Phuket, Thailand. Using 450 questionnaire responses from Muslim tourists who interacted with the platform or evaluated its core functions in a realistic usage scenario, the study integrated Unified Theory of Acceptance and Use of Technology (UTAUT)-based explanation with machine learning prediction. The analysis tested performance expectancy, effort expectancy, social influence, facilitating conditions, hedonic motivation, habit, and behavioral intention through measurement validation, structural equation modeling (SEM), regression, predictive benchmarking, ablation analysis, and explainability. Measurement results supported construct reliability, convergent validity, discriminant validity, and common-method bias diagnostics. However, the cross-sectional and self-reported nature of the data required cautious interpretation of the SEM paths. All six UTAUT constructs positively predicted behavioral intention, with habit, social influence, and hedonic motivation showing the strongest effects. Among all predictive configurations, the construct-only linear regression model achieved the best performance (root mean squared error (RMSE) = 0.251, R 2 = 0.861), outperforming item-level, hybrid, demographic, and nonlinear alternatives. Ablation findings indicated that removing demographic or expanded feature groups had a limited effect, whereas removing full UTAUT construct families substantially weakened performance, especially when habit and social influence were excluded. Feature-importance and SHapley Additive exPlanations (SHAP) analyses showed broad alignment with regression coefficients, supporting the value of theoretically grounded, parsimonious feature construction. The findings suggest that halal tourism platform developers should prioritize repeated-use support, social trust signals, enjoyable service experience, and practical usability rather than relying only on technical sophistication.

Received: 21 April 2026 | Revised: 12 June 2026 | Accepted: 14 July 2026

Conflicts of Interest

The authors declare that they have no conflicts of interest to this work.

Data Availability Statement

The data that support the findings of this study are openly available in GitHub at https://github.com/pjarupunphol/Datasets/blob/main/Halal.zip.

Author Contribution Statement

Nasith Laosen: Conceptualization, Methodology, Validation, Investigation, Resources, Data curation, Writing – original draft, Writing – review & editing, Supervision, Project administration, Funding acquisition. Tanagrit Chansaeng: Methodology, Software, Data curation, Writing – review & editing, Visualization. Wipawan Buathong: Investigation, Resources, Data curation, Writing – review & editing. Atipan Saimmai: Validation, Investigation, Resources, Writing – review & editing. Pita Jarupunphol: Conceptualization, Methodology, Software, Validation, Formal analysis, Data curation, Writing – original draft, Writing – review & editing, Visualization, Supervision, Project administration.