A recent study conducted by researchers from the Department of Software Engineering and the member of the MYDASS Research Group and the OnTheEdge R&D Project, at the University of Granada (UGR) introduces an innovative approach to predicting overnight stays in rural areas using AI. The authors, Daniel Bolaños-Martínez, María Bermúdez-Edo, and José Luis Garrido, developed a model based on License Plate Recognition (LPR) data to forecast tourist behavior in the Alpujarra region, combining vehicle movement data with contextual factors like holidays and socio-economic status.
Key Contributions
The study’s model classifies vehicle stays into three categories: day visits (vehicles stay in the area without overnighting), short visits (vehicles stay between one and five nights), and extended visits (vehicles remain for more than five nights). This classification allows for a more segmented and accurate prediction, helping local authorities make informed decisions. For instance, vehicles staying longer have a greater impact on the area in terms of both economic contribution and resource usage. This segmentation can guide policies, such as parking fees or infrastructure development, tailored to the type of visit.
The authors highlight the challenges faced by rural regions like the Alpujarra, where the combination of limited infrastructure and seasonal tourism demands precise management tools. Their model integrates LPR data with socio-economic variables to predict how long tourists will stay, enabling more accurate forecasting and decision-making. The study emphasizes the importance of combining multiple data sources, which is often neglected in traditional traffic studies. “Our approach promotes a more sustainable management of resources by reducing the use of unnecessary data and processing only the most relevant information,” explains Daniel Bolaños, lead author of the paper and one of the project’s researchers.
Results and Applications
The findings offer practical implications for rural tourism management. By analyzing vehicle movements and predicting stay durations, local governments can better allocate resources, plan for infrastructure needs, and optimize parking and road usage. This data-driven approach helps local authorities not only manage current traffic but also develop long-term sustainable tourism strategies in rural areas.
Collaborative Research
This study is part of a larger interdisciplinary project at UGR, involving the MYDASS and ISDE research groups. Supported by the EU’s LifeWatch ERIC initiative, the research analyzed vehicle movements in the Alpujarra region for over a year. The project reflects the power of collaboration between computer scientists, economists, and public authorities, demonstrating how interdisciplinary research can solve real-world challenges.
Reference
Bolaños-Martinez, D., Garrido, J.L. & Bermudez-Edo, M. Predicting overnights in smart villages: The importance of context information. Int. J. Mach. Learn. & Cyber. (2024). https://doi.org/10.1007/s13042-024-02337-7
