Developing Conversational AI to Enhance the Tourist Experience: A Methodological Framework Applied to the Case of Porto

Abstract

Tourism in Porto has expanded considerably over the last decade, with the number of visitors doubling between 2009 and 2019 and continuing to rise in recent years. This rapid growth creates challenges for the sustainable management of tourist inflows and highlights the need for innovative, data-driven solutions that can both support visitors and relieve pressure on local infrastructures. In response, this study proposes the development of a conversational AI system specifically designed to assist tourists in their daily activities and improve their overall experience of the city. The conversational AI is based on a Retrieval-Augmented Generation (RAG) framework, which combines an information retrieval component with a generative model to deliver accurate, context-aware responses. To ensure reliability, the system draws on more than 7,000 documents from diverse sources, including cultural guides, tourism platforms, and official city websites. Several experiments were conducted to identify the best performing system configuration, testing different retrieval strategies, ranking methods, and model architectures. The final solution demonstrates high retrieval accuracy and generates responses that score strongly on semantic similarity and answer quality metrics. Overall, the study demonstrates the potential of conversational AI systems as valuable tools for urban destinations facing growing tourist demand. Beyond Porto, this work illustrates how data-efficient conversational systems can support sustainable tourism management, improve the visitor experience, and serve as scalable solutions for cities with similar challenges worldwide.