Emotion Classification by Jointly Learning to Lexiconize and Classify

Deyu Zhou, Shuangzhi Wu, Qing Wang, Jun Xie, Zhaopeng Tu, Mu Li


Abstract
Emotion lexicons have been shown effective for emotion classification (Baziotis et al., 2018). Previous studies handle emotion lexicon construction and emotion classification separately. In this paper, we propose an emotional network (EmNet) to jointly learn sentence emotions and construct emotion lexicons which are dynamically adapted to a given context. The dynamic emotion lexicons are useful for handling words with multiple emotions based on different context, which can effectively improve the classification accuracy. We validate the approach on two representative architectures – LSTM and BERT, demonstrating its superiority on identifying emotions in Tweets. Our model outperforms several approaches proposed in previous studies and achieves new state-of-the-art on the benchmark Twitter dataset.
Anthology ID:
2020.coling-main.288
Volume:
Proceedings of the 28th International Conference on Computational Linguistics
Month:
December
Year:
2020
Address:
Barcelona, Spain (Online)
Editors:
Donia Scott, Nuria Bel, Chengqing Zong
Venue:
COLING
SIG:
Publisher:
International Committee on Computational Linguistics
Note:
Pages:
3235–3245
Language:
URL:
https://aclanthology.org/2020.coling-main.288
DOI:
10.18653/v1/2020.coling-main.288
Bibkey:
Cite (ACL):
Deyu Zhou, Shuangzhi Wu, Qing Wang, Jun Xie, Zhaopeng Tu, and Mu Li. 2020. Emotion Classification by Jointly Learning to Lexiconize and Classify. In Proceedings of the 28th International Conference on Computational Linguistics, pages 3235–3245, Barcelona, Spain (Online). International Committee on Computational Linguistics.
Cite (Informal):
Emotion Classification by Jointly Learning to Lexiconize and Classify (Zhou et al., COLING 2020)
Copy Citation:
PDF:
https://aclanthology.org/2020.coling-main.288.pdf