HiD-CATCH
History in Dialogue: A Collaborative Approach to Heritage Tours with AI Chatbots (HiD-CATCH). We are exploring how LLM-based chatbots can complement and augment the human tourist guide's expertise. The project is funded through the Research Excellence Programme 2024 by Xjenza Malta.
Funding
Research Excellence Programme
Year
2024-26
Category
Research & Development
Type of Work
Chat-Based Solutions
Large Language Models
Artificial Intelligence
Visitor Experience
The HiD-CATCH project explores the potential of Large Language Models (LLMs) to rethink the notion of historical walking tours, transforming them into immersive and interactive learning experiences. While acknowledging the risks of inaccuracies inherent in LLMs, the project aims to address these challenges proactively. Through a curated collaborative effort involving experts from diverse backgrounds, including heritage tourism practitioners, AI developers, and historians, we will develop and pilot an LLM-powered conversational system. Tourist interactions with historical figures, or even the location itself, provide opportunities for personalised knowledge exchange and deeper engagement with the past. The project employs a rigorous mixed-method approach to underpin both technological innovation and theoretical development.
Key outputs include an AI-Human Interaction Framework offering design principles for successful LLM integration, specialised protocols for training to ensure historical accuracy, and a toolkit for conversational analysis. Functional prototypes with iterative evaluation will inform the development of a Guide Training & Upskilling Manual. This project generates data for both theoretical advancements and practical guidelines, ensuring guides remain central to the experience and enhancing rather than replacing their expertise. Press outreach and research publications will extend the project's impact, establishing a strong model for the future of technologically-enhanced historical discovery that maintains a human-centred approach.
