Back to search

University of the Aegean

Generative computational storytelling for cultural heritage: contextual and adaptive narrative mechanisms

Abstract

dc:description

This doctoral thesis investigates the intersection of artificial intelligence, interactive narrative, and immersive technologies for enhancing user experiences in cultural heritage (CH) settings. Titled ”Generative computational storytelling for cultural heritage: contextual and adaptive narrative mechanisms”, the work introduces novel computational approaches to storytelling, emphasizing emergent, multimodal, and user-adaptive narratives. The research begins with an extensive survey of 92 computational storytelling projects and 37 augmented reality smart glasses (ARSG) applications, systematically categorized by purpose, media, and interaction type. It reveals a vibrant landscape of interactive digital narratives (IDN), particularly in CH, education, healthcare, and entertainment. Building on this foundation, the thesis proposes a series of innovative systems integrating tangible user interfaces (TUIs), digital twins, and large language models (LLMs). Notably, it introduces MAGICAL (Museum AI Guide for Augmenting Cultural heritage with Intelligent Language model), a hybrid museum guide and recommendation system powered by GPTbased language models. MAGICAL supports dynamic generation of personalized narratives and spatially aware recommendations, trained on contextual data such as exhibit location, chronology, and user role. Methodologically, the thesis combines laboratory and field experimentation. It details three field deployments of MAGICAL in a contemporary art exhibition, a fine arts museum, and an archaeological museum. Each scenario involved contextual training of GPT4, interface adaptation, and implementation of voice-based interactions. Evaluation included statistical analysis of user engagement and qualitative interviews, utilizing modified versions the User Experience Questionnaire (UEQ) and Museum Experience Scale (MES). The results indicate that MAGICAL significantly enhances perceived user engagement, personalization, and satisfaction. The system’s adaptability across different cultural domains underscores its generalizability and scientific impact.Concluding, the thesis outlines challenges in AI-based narrative systems—including hallucination, lack of spatial awareness, and ethical concerns—and proposes future research directions. These include integration of RAG (Retrieve-Augmented Generation), development of context-aware LLMs, and expansion of embodied storytelling systems using ARSGs and robotics. This work contributes a unified framework for intelligent storytelling systems in CH, demonstrating how AI can meaningfully transform visitor interaction and cultural engagement.

Degree

thesis:*
Grantor dc:publisher
University of the Aegean
Year dc:date
2025

Author and committee

dc:creator, dc:contributor.*
Authors dc:creator
  • Trichopoulos, Georgios
  • Τριχόπουλος, Γεώργιος

Subjects

dc:subject × 18

Rights

Language dc:language
eng

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/60187

Chain of custody

source
Harvested from
Greek National Archive of PhD Theses
Base URL
phdtheses.ekt.gr/eadd_oai/request
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
citation

Trichopoulos, Georgios; Τριχόπουλος, Γεώργιος. Generative computational storytelling for cultural heritage: contextual and adaptive narrative mechanisms. University of the Aegean, 2025. http://hdl.handle.net/10442/hedi/60187