Abstract
Itinerary recommendation is a complex sequence prediction problem with numerous practical applications. The task becomes significantly more challenging when optimizing multiple factors simultaneously, such as user queuing times, crowd levels, attraction popularity, walking durations, and operating hours. These factors, combined with the dynamic and unpredictable nature of visitor flow, introduce substantial complexities, particularly when accounting for collective user behavior. Existing solutions often adopt a single-user perspective, overlooking critical challenges arising from natural crowd dynamics. For example, the Selfish Routing problem illustrates how individual decision-making can lead to suboptimal outcomes for the group as a whole. To address these challenges, we propose the Strategic and Crowd-Aware Itinerary Recommendation (SCAIR) algorithm, which integrates real-world crowd behavior into route planning to optimize group utility. SCAIR models itinerary recommendation as a Markov Decision Process (MDP) and incorporates a novel State Encoding mechanism that facilitates real-time, efficient itinerary planning and resource allocation in linear time. By prioritizing group outcomes over individual preferences, SCAIR explicitly mitigates the adverse effects of selfish routing. We conduct extensive evaluations of SCAIR using a large-scale, real-world theme park dataset, benchmarking it against several competitive and realistic baselines. Our results demonstrate that SCAIR consistently outperforms these baselines, effectively addressing the limitations of selfish routing and significantly enhancing overall group utility across four major theme parks.