Abstract:
Tourism significantly boosts e-commerce by guiding tourists to points of interest
(POIs) among countless options online. Recommendation systems have
transformed the industry by delivering accurate, preference-based suggestions,
aiding in early identification and management. However, these systems often
struggle with robustness and personalization when limited by insufficient
personal data. In this paper presents the development of a Smart Tourism
Recommendation System that leverages deep learning to deliver personalized,
context-aware travel recommendations. The system employs a multi-stage
approach combining content-based filtering, Neural Collaborative Filtering
(NCF), and machine learning algorithms, specifically LightGBM and XGBoost,
alongside LightGCN—a graph-based deep learning model for collaborative
filtering
