Low-resource MobileBERT for emotion recognition in imbalanced text datasets mitigating challenges with limited resources.
Modern dialogue systems rely on emotion recognition in conversation (ERC) as a core element enabling empathetic and human-like interactions. However, the weak correlation between emotions and semantics poses significant challenges to emotion recognition in dialogue. Semantically similar utterances c...
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Public Library of Science (PLoS)
2025-01-01
|
| Series: | PLoS ONE |
| Online Access: | https://doi.org/10.1371/journal.pone.0312867 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832540197379637248 |
|---|---|
| author | Muhammad Hussain Caikou Chen Sami S Albouq Khlood Shinan Fatmah Alanazi Muhammad Waseem Iqbal M Usman Ashraf |
| author_facet | Muhammad Hussain Caikou Chen Sami S Albouq Khlood Shinan Fatmah Alanazi Muhammad Waseem Iqbal M Usman Ashraf |
| author_sort | Muhammad Hussain |
| collection | DOAJ |
| description | Modern dialogue systems rely on emotion recognition in conversation (ERC) as a core element enabling empathetic and human-like interactions. However, the weak correlation between emotions and semantics poses significant challenges to emotion recognition in dialogue. Semantically similar utterances can express different types of emotions, depending on the context or speaker. In order to tackle this challenge, our paper proposes a novel loss called Focal Weighted Loss (FWL) with adversarial training and the compact language model MobileBERT. Our proposed loss function handles the problem of imbalanced emotion classification through Focal Weighted Loss and adversarial training and does not require large batch sizes or more computational resources. Our approach has been employed on four text emotion recognition benchmark datasets, MELD, EmoryNLP, DailyDialog and IEMOCAP demonstrating competitive performance. Extensive experiments on these benchmark datasets validate the effectiveness of our proposed FWL with adversarial training. This enables more human-like interactions on digital platforms. Our approach shows its potential to deliver competitive performance under limited resource constraints, comparable to large language models. |
| format | Article |
| id | doaj-art-911d5c8224624f71b424cbd56afa4708 |
| institution | Kabale University |
| issn | 1932-6203 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | Public Library of Science (PLoS) |
| record_format | Article |
| series | PLoS ONE |
| spelling | doaj-art-911d5c8224624f71b424cbd56afa47082025-02-05T05:32:13ZengPublic Library of Science (PLoS)PLoS ONE1932-62032025-01-01201e031286710.1371/journal.pone.0312867Low-resource MobileBERT for emotion recognition in imbalanced text datasets mitigating challenges with limited resources.Muhammad HussainCaikou ChenSami S AlbouqKhlood ShinanFatmah AlanaziMuhammad Waseem IqbalM Usman AshrafModern dialogue systems rely on emotion recognition in conversation (ERC) as a core element enabling empathetic and human-like interactions. However, the weak correlation between emotions and semantics poses significant challenges to emotion recognition in dialogue. Semantically similar utterances can express different types of emotions, depending on the context or speaker. In order to tackle this challenge, our paper proposes a novel loss called Focal Weighted Loss (FWL) with adversarial training and the compact language model MobileBERT. Our proposed loss function handles the problem of imbalanced emotion classification through Focal Weighted Loss and adversarial training and does not require large batch sizes or more computational resources. Our approach has been employed on four text emotion recognition benchmark datasets, MELD, EmoryNLP, DailyDialog and IEMOCAP demonstrating competitive performance. Extensive experiments on these benchmark datasets validate the effectiveness of our proposed FWL with adversarial training. This enables more human-like interactions on digital platforms. Our approach shows its potential to deliver competitive performance under limited resource constraints, comparable to large language models.https://doi.org/10.1371/journal.pone.0312867 |
| spellingShingle | Muhammad Hussain Caikou Chen Sami S Albouq Khlood Shinan Fatmah Alanazi Muhammad Waseem Iqbal M Usman Ashraf Low-resource MobileBERT for emotion recognition in imbalanced text datasets mitigating challenges with limited resources. PLoS ONE |
| title | Low-resource MobileBERT for emotion recognition in imbalanced text datasets mitigating challenges with limited resources. |
| title_full | Low-resource MobileBERT for emotion recognition in imbalanced text datasets mitigating challenges with limited resources. |
| title_fullStr | Low-resource MobileBERT for emotion recognition in imbalanced text datasets mitigating challenges with limited resources. |
| title_full_unstemmed | Low-resource MobileBERT for emotion recognition in imbalanced text datasets mitigating challenges with limited resources. |
| title_short | Low-resource MobileBERT for emotion recognition in imbalanced text datasets mitigating challenges with limited resources. |
| title_sort | low resource mobilebert for emotion recognition in imbalanced text datasets mitigating challenges with limited resources |
| url | https://doi.org/10.1371/journal.pone.0312867 |
| work_keys_str_mv | AT muhammadhussain lowresourcemobilebertforemotionrecognitioninimbalancedtextdatasetsmitigatingchallengeswithlimitedresources AT caikouchen lowresourcemobilebertforemotionrecognitioninimbalancedtextdatasetsmitigatingchallengeswithlimitedresources AT samisalbouq lowresourcemobilebertforemotionrecognitioninimbalancedtextdatasetsmitigatingchallengeswithlimitedresources AT khloodshinan lowresourcemobilebertforemotionrecognitioninimbalancedtextdatasetsmitigatingchallengeswithlimitedresources AT fatmahalanazi lowresourcemobilebertforemotionrecognitioninimbalancedtextdatasetsmitigatingchallengeswithlimitedresources AT muhammadwaseemiqbal lowresourcemobilebertforemotionrecognitioninimbalancedtextdatasetsmitigatingchallengeswithlimitedresources AT musmanashraf lowresourcemobilebertforemotionrecognitioninimbalancedtextdatasetsmitigatingchallengeswithlimitedresources |