With the rapid advancement of information and communication technologies, smart tourism has become a crucial means for improving the quality of tourism services and enhancing economic efficiency in the tourism sector. This work proposes an analysis method based on the artificial neural network to predict tourist behavior patterns through big data analysis, thereby optimizing the allocation of tourism resources. The work begins by collecting various data types, including basic visitor information, consumption records, and satisfaction evaluations, from a well-known smart tourism destination as research samples. By carefully configuring and optimizing parameters such as learning rate, batch size, and optimizers, the work develops an efficient artificial neural network model. Experimental validation using real-world data demonstrates that the model excels across several performance metrics, including accuracy, recall, precision, and F1 score, and shows significant advantages over traditional statistical methods. In addition, the survey results show that users are highly satisfied with personalized service recommendations, resource optimization, and the overall user experience, with 75% of users expressing satisfaction. This work not only makes an academic contribution to the field of smart tourism but also demonstrates significant potential in improving tourism economic efficiency and enhancing the visitor experience.