Culturally-Aware Tourism Recommendation System Architecture Based on Gaze Tracking and Multilingual Web Access Pattern Analysis

抄録

Recommendation systems for tourism typically focus on individual preferences, overlooking cultural contexts. We propose a culturally-aware system integrating gaze tracking with JNTO (Japan National Tourism Organization) data analysis across nine countries. Key findings: dramatic individualization (China 43.8%→96.0%) and regional specialization (Korean hot springs: 22-33% nationally vs. 68.3% in Kagoshima). Our three-layer framework combines gaze-based extraction with hierarchical cultural adaptation. Evaluation metrics balance cultural relevance (60-75%) with discovery (20-35%).