Enhancing Patient‐Centric Care: A Location‐Aware Hospital Recommender System Optimized by the Water Cycle Algorithm

ABSTRACT Healthcare recommender systems (HRSs) play a vital role in modern healthcare services, particularly during the COVID‐19 pandemic. These systems assist in identifying suitable hospitals for patients by prioritizing their needs and preferences. This article presents the development of a personalized, location‐based hospital recommender system designed for individuals seeking healthcare centers during the COVID‐19 crisis. The initial phase involves collecting public participation data (PPD), patients' personal information, and hospital attributes to identify potential hospitals and build a knowledge‐based recommendation framework. A total of 166 participants were used for user clustering in the personalization stage. In the next phase, the system integrates users' location, visit time, and preferences to deliver personalized recommendations tailored to their needs and geographical context. To further improve recommendation quality, the system employs the water cycle algorithm (WCA), a metaheuristic optimization method that enhances hospital selection by maximizing accessibility to public transportation stations (PTSs), pharmacies, and hospital bed availability, while minimizing travel time. The novelty of the proposed system lies in integrating personalized user characteristics and preferences, spatial accessibility modeling, real‐world hospital capacity constraints, and metaheuristic optimization within a unified framework. The system was implemented in Tehran, Iran. Performance was evaluated using 20 users across multiple scenarios, producing 50 recommendation outcomes. Results demonstrate that the WCA‐based system achieved an F1 score of 84.93%, an AUC of 0.727, an average runtime of 5.23 s, and a user satisfaction (CSAT) score of 74.67%, indicating high diversity, efficient performance, and notable patient satisfaction.