Fine-Grained Emotion Detection on GoEmotions: Experimental Comparison of Classical Machine Learning, BiLSTM, and Transformer Models
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2026
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866910000853549056 |
|---|---|
| author | Harutyunyan, Ani Kumar, Sachin |
| author_facet | Harutyunyan, Ani Kumar, Sachin |
| contents | Fine-grained emotion recognition is a challenging multi-label NLP task due to label overlap and class imbalance. In this work, we benchmark three modeling families on the GoEmotions dataset: a TF-IDF-based logistic regression system trained with binary relevance, a BiLSTM with attention, and a BERT model fine-tuned for multi-label classification. Experiments follow the official train/validation/test split, and imbalance is mitigated using inverse-frequency class weights. Across several metrics, namely Micro-F1, Macro-F1, Hamming Loss, and Subset Accuracy, we observe that logistic regression attains the highest Micro-F1 of 0.51, while BERT achieves the best overall balance surpassing the official paper's reported results, reaching Macro-F1 0.49, Hamming Loss 0.036, and Subset Accuracy 0.36. This suggests that frequent emotions often rely on surface lexical cues, whereas contextual representations improve performance on rarer emotions and more ambiguous examples. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_18162 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Fine-Grained Emotion Detection on GoEmotions: Experimental Comparison of Classical Machine Learning, BiLSTM, and Transformer Models Harutyunyan, Ani Kumar, Sachin Computation and Language Fine-grained emotion recognition is a challenging multi-label NLP task due to label overlap and class imbalance. In this work, we benchmark three modeling families on the GoEmotions dataset: a TF-IDF-based logistic regression system trained with binary relevance, a BiLSTM with attention, and a BERT model fine-tuned for multi-label classification. Experiments follow the official train/validation/test split, and imbalance is mitigated using inverse-frequency class weights. Across several metrics, namely Micro-F1, Macro-F1, Hamming Loss, and Subset Accuracy, we observe that logistic regression attains the highest Micro-F1 of 0.51, while BERT achieves the best overall balance surpassing the official paper's reported results, reaching Macro-F1 0.49, Hamming Loss 0.036, and Subset Accuracy 0.36. This suggests that frequent emotions often rely on surface lexical cues, whereas contextual representations improve performance on rarer emotions and more ambiguous examples. |
| title | Fine-Grained Emotion Detection on GoEmotions: Experimental Comparison of Classical Machine Learning, BiLSTM, and Transformer Models |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2601.18162 |