Explore for a day? Generating personalized itineraries that fit spatial heterogeneity of tourist attractions

Abstract Recommender systems are widely adopted by firms as an innovative personalization tool across various industries. Most of the existing tour recommender systems treat the spatial structure of tourist attractions as a single type, which neglects the spatial heterogeneity among these attractions. This study attempts to address this problem by modeling the spatial heterogeneity in the design of personalized trips. We propose a two-phase heuristic approach, which involves an improved artificial bee colony algorithm and a differential evolution algorithm. The results of a field experiment confirm that our new model outperforms the benchmark models in maximizing customer utilities.