DeBERTa-v3-large achieves the highest mean F1 (0.662) among compared models for binary detection of eight Plutchik emotions in Japanese WRIME posts, though the paper's stated accuracy advantage is not supported by its own table.
In: Proceedings of the Workshop on Computational Ling uistics and Clinical Psychology: From Linguistic Signal to Clinical Reality, pp
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Performance Evaluation of Emotion Classification in Japanese Using RoBERTa and DeBERTa
DeBERTa-v3-large achieves the highest mean F1 (0.662) among compared models for binary detection of eight Plutchik emotions in Japanese WRIME posts, though the paper's stated accuracy advantage is not supported by its own table.