Elamur Research⌕
☰
home / papers / 2019 / Multi-Task Stance Detection with Sentiment and Stance Lexicons

Multi-Task Stance Detection with Sentiment and Stance Lexicons

Summary

Stance detection asks whether the author of a text supports, opposes, or is neutral toward a given target, which often goes unnamed in the text itself. Li and Caragea address this for tweets in SemEval-2016 Task 6.A, which covers five targets such as Atheism and Hillary Clinton. Their model, AT-JSS-Lex, encodes fastText embeddings with a BiLSTM and trains two tasks jointly: three-way sentiment classification as an auxiliary objective and stance detection as the main one. The stance branch joins two signals: a target-aware attention layer that receives the target's word vector, and the output of the sentiment attention layer. An extra loss term pushes the normalized attention weights toward words flagged by two lexicons. One is the Hu and Liu (2004) sentiment list. The other is a new stance lexicon of roughly 2,000 target-related words, compiled by hand from training tweets plus tweets gathered via hashtags. The system achieves the best macro and micro F1 among the compared deep models. This suggests that sentiment supervision and lexicon-guided attention help when labelled stance data is scarce.

Key points

  • Data: SemEval-2016 stance dataset with 2,914 training and 1,249 test tweets over five targets. Each tweet carries a stance label (Favor/Against/None) and a sentiment label.
  • Stance lexicon: about 400 words per target, built manually from the training data plus about 1,000 extra tweets per target collected via hashtags such as #Hillary2016. It is released at github.com/chuchun8/EMNLP19-Stance.
  • Loss settings: main-task weight λ = 0.7 and lexicon-loss weight β = 0.025. The lexicon term is the squared gap between the max-normalized summed attention and a binary lexicon vector.
  • Results: AT-JSS-Lex scores 65.33 MacF_avg and 72.33 MicF_avg. It beats the JOINT model by 5.17 and 3.11 points and leads on every target except Atheism.
  • Ablation and errors: removing the lexicon loss costs 0.75 (macro) and 0.27 (micro) points. The main remaining error sources are mislabeled tweets and unsplit compound hashtags like #NoHillary.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

Fields: Linguistics & languages

Cited by

33 papers of our library cite it (of the 140673 whose bibliographies we read); 10 of them have a page here, 23 not yet.

Cite

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/d19-1657

@inproceedings{li2019multi,
  title = {{Multi-Task Stance Detection with Sentiment and Stance Lexicons}},
  author = {Li, Yingjie and Caragea, Cornelia},
  booktitle = {Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP)},
  year = {2019},
  doi = {10.18653/v1/d19-1657},
  url = {https://doi.org/10.18653/v1/d19-1657}
}

References

24 references, 4 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Dzmitry Bahdanau, Kyunghyun Cho, and Yoshua Bengio. 2015. Neural machine translation by jointly learning to align and translate. In 3rd International Conference on Learning Representations. find our card
  2. Georgios Balikas, Simon Moura, and Massih-Reza Amini. 2017. Multitask learning for fine-grained twitter sentiment analysis. In Proceedings of the 40th international ACM SIGIR conference on research and development in information retrieval, pages 1005–1008. find
  3. Piotr Bojanowski, Edouard Grave, Armand Joulin, and Tomas Mikolov. 2017. Enriching word vectors with subword information. Transactions of the Association for Computational Linguistics, 5:135–146. find
  4. Arman Cohan, Waleed Ammar, Madeleine van Zuylen, and Field Cady. 2019. Structural scaffolds for citation intent classification in scientific publications. arXiv preprint arXiv:1904.01608. find
  5. Jiachen Du, Ruifeng Xu, Yulan He, and Lin Gui. 2017. Stance classification with target-specific neural attention networks. In Proceedings of the 26th International Joint Conference on Artificial Intelligence, pages 3988–3994. find
  6. Sepp Hochreiter and Jürgen Schmidhuber. 1997. Long short-term memory. Neural computation, 9(8):1735–1780. find
  7. Minqing Hu and Bing Liu. 2004. Mining and summarizing customer reviews. In Proceedings of the tenth ACM SIGKDD international conference on Knowledge discovery and data mining, pages 168–177. find
  8. Diederik P. Kingma and Jimmy Ba. 2015. Adam: A method for stochastic optimization. In 3rd International Conference on Learning Representations. find
  9. Pengfei Liu, Xipeng Qiu, and Xuanjing Huang. 2016. Recurrent neural network for text classification with multi-task learning. In Proceedings of the Twenty-Fifth International Joint Conference on Artificial Intelligence, pages 2873–2879. find
  10. Pengfei Liu, Xipeng Qiu, and Xuanjing Huang. 2017. Adversarial multi-task learning for text classification. In Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 1–10. find
  11. Tomas Mikolov, Ilya Sutskever, Kai Chen, Greg S Corrado, and Jeff Dean. 2013. Distributed representations of words and phrases and their compositionality. In Advances in neural information processing systems, pages 3111–3119. find
  12. Saif Mohammad, Svetlana Kiritchenko, Parinaz Sobhani, Xiao-Dan Zhu, and Colin Cherry. 2016a. A dataset for detecting stance in tweets. In LREC. find
  13. Saif Mohammad, Svetlana Kiritchenko, Parinaz Sobhani, Xiaodan Zhu, and Colin Cherry. 2016b. Semeval-2016 task 6: Detecting stance in tweets. In Proceedings of the 10th International Workshop on Semantic Evaluation (SemEval-2016), pages 31–41. find
  14. Saif M Mohammad, Parinaz Sobhani, and Svetlana Kiritchenko. 2017. Stance and sentiment in tweets. ACM Transactions on Internet Technology (TOIT), 17(3):26. find
  15. Mike Schuster and Kuldip K Paliwal. 1997. Bidirectional recurrent neural networks. IEEE Transactions on Signal Processing, 45(11):2673–2681. find
  16. Parinaz Sobhani, Saif Mohammad, and Svetlana Kiritchenko. 2016. Detecting stance in tweets and analyzing its interaction with sentiment. In Proceedings of the Fifth Joint Conference on Lexical and Computational Semantics, pages 159–169. find
  17. Qingying Sun, Zhongqing Wang, Shoushan Li, Qiaoming Zhu, and Guodong Zhou. 2019. Stance detection via sentiment information and neural network model. Frontiers of Computer Science, 13(1):127– 138. find
  18. Qingying Sun, Zhongqing Wang, Qiaoming Zhu, and Guodong Zhou. 2018. Stance detection with hierarchical attention network. In Proceedings of the 27th International Conference on Computational Linguistics, pages 2399–2409. find our card
  19. Prashanth Vijayaraghavan, Ivan Sysoev, Soroush Vosoughi, and Deb Roy. 2016. DeepStance at SemEval-2016 task 6: Detecting stance in tweets using character and word-level CNNs. In Proceedings of the 10th International Workshop on Semantic Evaluation (SemEval-2016), pages 413–419. find
  20. Penghui Wei, Wenji Mao, and Daniel Zeng. 2018. A target-guided neural memory model for stance detection in twitter. In 2018 International Joint Conference on Neural Networks (IJCNN), pages 1–8. find
  21. Wan Wei, Xiao Zhang, Xuqin Liu, Wei Chen, and Tengjiao Wang. 2016. pkudblab at semeval-2016 task 6: A specific convolutional neural network system for effective stance detection. In Proceedings of the 10th International Workshop on Semantic Evaluation (SemEval-2016), pages 384–388. find our card
  22. Jianfei Yu, Luis Marujo, Jing Jiang, Pradeep Karu-turi, and William Brendel. 2018. Improving multi-label emotion classification via sentiment classification with dual attention transfer network. In Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing, pages 1097– 11… find
  23. Guido Zarrella and Amy Marsh. 2016. MITRE at SemEval-2016 task 6: Transfer learning for stance detection. In Proceedings of the 10th International Workshop on Semantic Evaluation (SemEval-2016), pages 458–463. find our card
  24. Yiwei Zhou, Alexandra I Cristea, and Lei Shi. 2017. Connecting targets to tweets: Semantic attention-based model for target-specific stance detection. In International Conference on Web Information Systems Engineering, pages 18–32. find

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/li-2019-multi-task-stance-detection-sentiment/