arbiter
JournalsPricingAccount

An Interpretable and Visual Deep Learning Framework for Korean Sentiment Analysis in Multi-Domain Service Scenarios

Ping Zhao

International Journal of Information Systems in the Service Sector2025https://doi.org/10.4018/ijisss.396817article
AJG 1
Weight
0.50

What the paper says

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.

Open paper page →

Cite this paper

https://doi.org/https://doi.org/10.4018/ijisss.396817

Or copy a formatted citation

@article{ping2025,
  title        = {{An Interpretable and Visual Deep Learning Framework for Korean Sentiment Analysis in Multi-Domain Service Scenarios}},
  author       = {Ping Zhao},
  journal      = {International Journal of Information Systems in the Service Sector},
  year         = {2025},
  doi          = {https://doi.org/https://doi.org/10.4018/ijisss.396817},
}

Paste directly into BibTeX, Zotero, or your reference manager.

Flag this paper

An Interpretable and Visual Deep Learning Framework for Korean Sentiment Analysis in Multi-Domain Service Scenarios

Flags are reviewed by the Arbiter methodology team within 5 business days.

Evidence weight

0.50

Balanced mode · F 0.40 / M 0.15 / V 0.05 / R 0.40

F · citation impact0.50 × 0.4 = 0.20
M · momentum0.50 × 0.15 = 0.07
V · venue signal0.50 × 0.05 = 0.03
R · text relevance †0.50 × 0.4 = 0.20

† Text relevance is estimated at 0.50 on the detail page — for your query’s actual relevance score, open this paper from a search result.