An Interpretable and Visual Deep Learning Framework for Korean Sentiment Analysis in Multi-Domain Service Scenarios
International Journal of Information Systems in the Service Sector · 23.12.2025 · 16 (1) · 1-17
DOI: 10.4018/ijisss.396817 ↗Kurz gefasst
Fachartikel in der Zeitschrift International Journal of Information Systems in the Service Sector, erschienen am 23.12.2025. Verfasst von Ping Zhao. Die Arbeit nennt 32 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 3.0).
Abstract (Originalsprache)
Sentiment analysis is vital for actionable insights in service sectors, but Korean sentiment analysis faces challenges from the language's agglutinative morphology, honorifics, code-mixing, and limited interpretable models. This study proposes an “interpretable-visible-migratable” deep learning framework for multi-domain Korean text in service scenarios, integrating subword preprocessing with feedback, a dual-channel BiGRU-Att for joint sentiment polarity-intensity optimization, and a multi-layer explainability system. A 30,000-sentence FAIR-compliant Korean dataset is also released. Experiments show 90.9% polarity accuracy, 0.38 intensity MAE, and 8.6/10 transparency satisfaction, balancing performance and interpretability to enhance sentiment analysis applicability in service sector information systems.
Zitieren
APA
Zhao, P. (2025). An Interpretable and Visual Deep Learning Framework for Korean Sentiment Analysis in Multi-Domain Service Scenarios. International Journal of Information Systems in the Service Sector, 16(1), 1-17. https://doi.org/10.4018/ijisss.396817
Vancouver
Zhao P. An Interpretable and Visual Deep Learning Framework for Korean Sentiment Analysis in Multi-Domain Service Scenarios. International Journal of Information Systems in the Service Sector. 2025;16(1):1-17. doi:10.4018/ijisss.396817
BibTeX
@article{Zhao2025,
author = {Zhao, Ping},
title = {An Interpretable and Visual Deep Learning Framework for Korean Sentiment Analysis in Multi-Domain Service Scenarios},
journal = {International Journal of Information Systems in the Service Sector},
year = {2025},
volume = {16},
number = {1},
pages = {1--17},
doi = {10.4018/ijisss.396817}
}
EUVIMED