A Scalable Deep Reinforcement Learning Framework for Enhancing Intangible Cultural Heritage Digital Protection
International Journal of e-Collaboration · 11.11.2026 · 22 (1) · 1-43
DOI: 10.4018/ijec.427453 ↗Kurz gefasst
Fachartikel in der Zeitschrift International Journal of e-Collaboration, erschienen am 11.11.2026. Verfasst von Dongjin Mao. Die Arbeit nennt 35 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 3.0).
Abstract (Originalsprache)
This study designs a Deep Reinforcement Learning (DRL)-based optimization framework to improve digital protection strategies of ICH in China. This proposed approach uses the cultural heritage digital preservation data with 10000 records obtained and the data is preprocessed by several techniques such as data cleaning by removing duplicate records, missing values, noise reduction by removing stop words, Wiener filter, and normalizing the data by min-max normalization. TF-IDF is used for text features, CNN for visual features, and MFCC for audio features. The DRL-based Dynamic Differential Annealed Optimized Advantage Actor-Critic (DDA-A2C) method is designed to dynamically learn optimal preservation actions, including the prioritization of digitization, allocation of resources, and other cultural domains. The DDA optimization fine tunes the DDA parameters. The results show that the DDA-A2C improves accuracy to 98.92%. Findings highlight the potential of AI-driven strategies to support sustainable, large-scale digital preservation and promotion of China's intangible cultural heritage.
Zitieren
APA
Mao, D. (2026). A Scalable Deep Reinforcement Learning Framework for Enhancing Intangible Cultural Heritage Digital Protection. International Journal of e-Collaboration, 22(1), 1-43. https://doi.org/10.4018/ijec.427453
Vancouver
Mao D. A Scalable Deep Reinforcement Learning Framework for Enhancing Intangible Cultural Heritage Digital Protection. International Journal of e-Collaboration. 2026;22(1):1-43. doi:10.4018/ijec.427453
BibTeX
@article{Mao2026,
author = {Mao, Dongjin},
title = {A Scalable Deep Reinforcement Learning Framework for Enhancing Intangible Cultural Heritage Digital Protection},
journal = {International Journal of e-Collaboration},
year = {2026},
volume = {22},
number = {1},
pages = {1--43},
doi = {10.4018/ijec.427453}
}
EUVIMED