Papers by Tran Quang Hung
ViGoEmotions: A Benchmark Dataset For Fine-grained Emotion Detection on Vietnamese Texts (2026.eacl-long)
Copied to clipboard
| Challenge: | Recent advances in NLP have greatly improved outcomes in emotion prediction and harmful content detection. |
| Approach: | They propose to classify Vietnamese comments into 27 distinct emotions using a model-based lexical normalization system and a transformer-based model. |
| Outcome: | The proposed corpus of 20,664 social media comments is based on a novel model that can support multiple architectures, but its quality and preprocessing strategies remain key factors influencing performance. |