Design and Development of a Rural Tourism Marketing System Using Deep Learning

This work aims to explore the optimization effects of marketing in the field of rural tourism with the support of artificial intelligence technologies such as deep learning. It first conducts a framework analysis of the requirements of the rural tourism marketing system, focusing on the recommendation system module. Subsequently, the RippleNet network is introduced, incorporating a scenic knowledge graph into the recommendation model. Simultaneously, the Spatiotemporal Graph Convolutional Attention Network (STGCAN) algorithm is introduced to build a rural tourism recommendation system based on RippleNet integrating STGCAN. Experimental results demonstrate that the system has achieved significant success in terms of loss value, prediction accuracy, recall, and F1 value, with an accuracy of 92.64%, recall of 89.65%, and F1 value of 92.12%, surpassing baseline algorithms such as Convolutional Neural Network (CNN). Additionally, the runtime of the proposed algorithm is significantly lower than that of other models, with a runtime of only 18.84 seconds for a data volume of 3000 records. Therefore, the proposed rural tourism recommendation system exhibits superior predictive performance, providing robust support for the optimization of marketing strategies in the tourism industry.