Interest-aware and context-adaptive model for personalised travel route recommendation

With the growing demand for personalised tourism, route recommendation has become a key issue in intelligent travel.Existing methods face limitations in personalisation, temporal rhythm, and adaptability to dynamic environments.We propose an interest-aware and context-adaptive route recommendation model (ICRR).First, an interest-aware adaptive attention mechanism integrates user interest vectors into graph attention networks to enable personalised representations.Second, a temporal segmentation optimiser leverages LSTM and attention to capture temporal dependencies and solve the orienteering problem with time constraints, using adaptive perturbation search to avoid local optima.Finally, a dynamic route refinement mechanism models environmental factors through reinforcement learning for real-time route adjustment.Experiments show that ICRR outperforms baselines in user satisfaction, recommendation accuracy, and robustness, offering an efficient solution for smart tourism and intelligent transportation.