Abstract
In the last several decades, cities all over the globe have seen the arrival of millions of tourists. Due to the rapid increase, innovative approaches are required to control tourist demand, generate precise forecasts, and offer tailored suggestions. This study presents the Urban Tourism Evaluation System (UTES), a novel hybrid method that combines collaborative filtering with cutting-edge techniques such as LSTM neural networks, fuzzy logic, big data analysis, and K-means clustering. The UTES system processes and analyzes large tourist datasets via its interconnected parts. At its core is a robust big data storage system that absorbs large amounts of tourist data. Tourists are categorized using K-means clustering algorithms according to their preferences and actions. Unclear inputs can be correctly deciphered by fuzzy logic. Long short-term memory (LSTM) networks allow for accurate demand forecasting in tourism by recording complex relationships between actual demand and predictive variables. Planners can use these projections to monitor and manage the number of tourists visiting cities. Simultaneously, the collaborative filtering tool analyzes traveller preferences using data mining, enabling highly personalized recommendations for locations, attractions, and activities. With its integrated components for visitor flow management and personalized recommendations, UTES is a comprehensive tourism management system that can evaluate data in real-time and respond with judgments. UTES revolutionizes the industry with data mining, neural networks, recommendation algorithms, big data analytics, and urban tourism management. Ultimately, UTES empowers data-driven decision-making, optimal visitor flow management, and highly personalized tourist experiences, meeting the diverse needs of modern urban travellers. The experimental results show that UTES achieves a demand forecasting accuracy of 93.4% and a recommendation accuracy of 96.7%, outperforming existing models. Additionally, UTES demonstrated lower MAE (78.2%) and RMSE (71.8%), superior real-time adaptability, clustering efficiency, and robust data uncertainty handling using fuzzy logic. These advantages make UTES a highly effective solution for optimizing urban tourism management.