Sistema de Recomendación Turística : Sistema de Recomendación Personalizada de Restaurantes.

Abstract

This document presents a restaurant recommendation system designed to enhance decision-making for tourists by integrating their personal preferences and geographic location. The system employs a hybrid approach, combining collaborative filtering and content-based techniques to deliver restaurant suggestions that best suit each user’s tastes and preferences. The recommendations are generated from Google’s top-rated restaurants, which are then ranked through sentiment analysis applied to user reviews. This sentiment analysis ensures that the suggested restaurants are highly relevant and customized to meet the tourists’ expectations. The restaurant data is obtained from the Google API, which guarantees constantly updated information and ensures that the suggested establishments are open and available. The system offers additional functionalities to enhance the user experience. For instance, it allows users to adjust the search radius based on their location, refine their options by cuisine type, price range, or Google ratings, and access the system via social media accounts. Additionally, users can view detailed information about restaurants and generate walking or driving routes to reach their destination, making the dining experience much more convenient and enjoyable for tourists. The system's development followed the Scrum framework, which enabled the team to efficiently respond to changing requirements and user feedback throughout the process. From a technical perspective, Angular and Express were employed to build the system’s infrastructure, ensuring a smooth and efficient user interface. Furthermore, the IBM Watson API was integrated to perform sentiment analysis, enhancing the system’s precision and improving the overall quality of recommendations. Throughout the project, the focus remained on optimizing user interaction, ensuring real-time data accuracy, and offering tailored recommendations that align with both individual preferences and current location.