A Dynamic Reconstruction System for Historical Museum Spatial Narratives Based on Generative Adversarial Networks (GANs)—A Semantic Coherence Protection Mechanism for Dispersed Cultural Relics

Authors

  • Rui Cheng International College, Krirk University, Bangkok, 10220, Thailand
  • Ren Zhou International College, Krirk University, Bangkok, 10220, Thailand

DOI:

https://doi.org/10.70917/ijcisim-2025-0182

Keywords:

TF-IDF algorithm; FCM clustering algorithm; self-attention mechanism; GAN; historical museum; spatial narrative

Abstract

 Historical museums offer dual insights into the revitalisation and creative dissemination of intangible cultural heritage, dynamically reconstructing historical memory through spatial narratives and fostering contemporary dialogue through the activation of historical spatial narratives. This article takes the dynamic reconstruction mechanism of spatial narratives in historical museums as its starting point, analysing the multi-layered narrative design of historical museums based on narrative theory and spatial narrative concepts, and proposing a semantic contextual reconstruction method for spatial narratives in historical museums. Based on this, a visual symbol extraction method for historical museums is constructed using filters and CNN, with TF-IDF algorithms to extract semantic information from visual symbols, and FCM clustering algorithms to perform keyword clustering of cultural relic space semantics. Self-attention mechanisms, spectral normalisation, and gradient normalisation are introduced to optimise GAN, constructing an improved GAN-based visual symbol translation model for historical museums. Taking the Beijing History Museum as the research object, data analysis was conducted on the model's performance and spatial narrative expression. The semantic space of historical museum artefacts was primarily divided into nine clustering labels, and the overall performance of the visual symbol translation model was satisfactory. After implementing the spatial narrative optimisation strategy, the participants' experience scores increased by 61.69%, and over 80% of the audience expressed approval for the historical museum's spatial narrative update mechanism. Therefore, by dynamically reconstructing the spatial narrative of historical museums, discrete artefacts can be given a more coherent preservation approach, thereby enhancing audience experience in historical museums and creating a more vibrant cultural emblem

Downloads

Download data is not yet available.

Downloads

Published

2025-12-21

How to Cite

Rui Cheng, & Ren Zhou. (2025). A Dynamic Reconstruction System for Historical Museum Spatial Narratives Based on Generative Adversarial Networks (GANs)—A Semantic Coherence Protection Mechanism for Dispersed Cultural Relics. International Journal of Computer Information Systems and Industrial Management Applications, 17, 18. https://doi.org/10.70917/ijcisim-2025-0182

Issue

Section

Original Articles