The Analysis of Tourist Satisfaction Integrating the Artistic Intelligence Convolutional Neural Network and Internet of Things Technology

Abstract

Aiming at the limitations of existing fire detection technology and the multi-dimensional challenge of tourist satisfaction analysis, this study proposes a series of innovative methods and models. Regarding fire detection, Depth Separable Convolution (DSC) and multi-scale detection structure are introduced to improve the You Only Look Once version 3 (YOLOv3) model. Moreover, the DSC-Anchor-Isoft-Non Maximum Suppression-YOLO (DAI-YOLO) model is implemented for the fire detection of scenic spots. The experimental results show that the precision, recall, and average precision of the DAI-YOLO model are 92.1%, 84.2%, and 84.6%, respectively, compared with other models, a minimum increase of 4.1%, 8.8%, and 6.0%, with higher detection accuracy and performance. Based on the analysis of tourist satisfaction, a comprehensive index system is constructed using the grounded theory of text mining, and emotion analysis is integrated into the satisfaction evaluation to reveal tourists' evaluation of scenic spots more comprehensively. According to the analysis, the environmental factor receives the highest satisfaction rating, reaching a positive rate of 98.78%. However, the satisfaction evaluation of scenic spot management is relatively low, accounting for 6.06% of the negative evaluation. The importance-satisfaction analysis reveals that the key factors affecting tourist satisfaction are traffic level, scenic spot tickets, and service. The results of this study provide valuable reference for the managers and researchers of scenic spots and are expected to contribute to the construction of a safer and more satisfying tourism experience.