Elamur Research⌕
☰
home / papers / 2021 / Learning Span-Level Interactions for Aspect Sentiment Triplet Extraction

Learning Span-Level Interactions for Aspect Sentiment Triplet Extraction

Paper · 2021 · Lu Xu, Yew Ken Chia and Lidong Bing

Summary

Aspect Sentiment Triplet Extraction (ASTE) asks a system to find, in a single pass, an aspect target, the opinion term about it, and the resulting sentiment. Earlier end-to-end systems scored target–opinion links word by word. This breaks down when either element spans several words, which happens in about a third of benchmark triplets. The authors introduce Span-ASTE, which enumerates candidate spans up to a maximum length and represents each one by its boundary tokens plus a learned width embedding. It then classifies the sentiment relation between whole target spans and whole opinion spans. Scoring all span pairs would cost O(n⁴), so a mention module keeps two separate shortlists, one for targets and one for opinions. This module is supervised with labels from the subtasks of identifying aspect terms (ATE) and identifying opinion terms (OTE). Evaluated on four benchmark datasets with BiLSTM and BERT encoders, the model beats prior methods on ASTE, ATE and OTE. The largest gains come on multi-word triplets. Sentiment stays consistent across full phrases, though single-word triplets still score noticeably higher than multi-word ones.

Key points

  • Architecture: three modules (sentence encoding, mention module, triplet module), with code released at github.com/chiayewken/Span-ASTE.
  • Encoders: a BiLSTM over 300-dimensional GloVe embeddings, or BERT with mean pooling over word pieces.
  • Pruning: a "dual-channel" strategy, guided by ATE/OTE supervision, ranks spans by their target and opinion mention scores. It keeps a pool of each type sized in proportion to sentence length, rather than one mixed pool as in Luan et al. (2019) and Wadden et al. (2019).
  • Ablation (average dev F1 over 4 datasets, 5 runs):
    • The full model scores 67.69.
    • Removing the width and distance embeddings costs 1.24.
    • Replacing boundary concatenation with max pooling costs 1.60, and with mean pooling 1.53.
  • Sentiment labels: Positive, Negative and Neutral, plus an Invalid class for target–opinion pairs with no relation.

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

 Institutions

🏢Alibaba GroupHangzhou · China64 works · 14,113 citations🎓Singapore University of Technology and DesignSingapore · Singapore33 works · 10,760 citations🏢Alibaba DAMO AcademyHangzhou · China28 works · 6,115 citations🏢Alibaba Group (Cayman Islands)George Town · Cayman Islands15 works · 3,209 citations

📚 Shelves 4The field of science this work belongs to (OpenAlex fields).Computer Science · 9,086 worksArtificial Intelligence · 8,889 worksArts and Humanities · 9,296 worksLanguage and Linguistics · 9,206 works

🧭 Topics 1Research subjects and languages the work is about.Sentiment & emotion · 608 works

Cite

Cite as Lu Xu, Yew Ken Chia and Lidong Bing. 2021. Learning Span-Level Interactions for Aspect Sentiment Triplet Extraction. Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers).

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2021.acl-long.367

@inproceedings{xu2021learning,
  title = {{Learning Span-Level Interactions for Aspect Sentiment Triplet Extraction}},
  author = {Lu Xu, Yew Ken Chia and Lidong Bing},
  booktitle = {Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers)},
  year = {2021},
  doi = {10.18653/v1/2021.acl-long.367},
  url = {https://doi.org/10.18653/v1/2021.acl-long.367}
}

References

38 references, 16 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Zhenzhong Lan, Mingda Chen, Sebastian Goodman, Kevin Gimpel, Piyush Sharma, and Radu Soricut. 2020. Albert: A lite bert for self-supervised learning of language representations. In Proc. of ICLR. find
  2. Peng Chen, Zhongqian Sun, Lidong Bing, and Wei Yang. 2017. Recurrent attention network on memory for aspect sentiment analysis. In Proc. of EMNLP.
  3. Kenton Lee, Luheng He, Mike Lewis, and Luke Zettlemoyer. 2017. End-to-end neural coreference resolution. In Proc. of EMNLP.
  4. Zhuang Chen and Tieyun Qian. 2020. Relation-aware collaborative learning for unified aspect-based sentiment analysis. In Proc. of ACL.
  5. Hao Li and Wei Lu. 2017. Learning latent sentiment scopes for entity-level sentiment analysis. In Proc. of AAAI. find
  6. Hongliang Dai and Yangqiu Song. 2019. Neural aspect and opinion term extraction with mined rules as weak supervision. In Proc. of ACL. find
  7. Hao Li and Wei Lu. 2019. Learning explicit and implicit structures for targeted sentiment analysis. In Proc. of EMNLP. find
  8. Xin Li, Lidong Bing, Wai Lam, and Bei Shi. 2018a. Transformation networks for target-oriented sentiment classification. In Proc. of ACL.
  9. Xin Li, Lidong Bing, Piji Li, and Wai Lam. 2019. A unified model for opinion target extraction and target sentiment prediction. In Proc. of AAAI. find
  10. Xin Li, Lidong Bing, Piji Li, Wai Lam, and Zhimou Yang. 2018b. Aspect term extraction with history attention and selective transformation. In Proc. of IJCAI. find
  11. Colin Raffel, Noam Shazeer, Adam Roberts, Katherine Lee, Sharan Narang, Michael Matena, Yanqi Zhou, Wei Li, and Peter J. Liu. 2020. Exploring the limits of transfer learning with a unified text-to-text transformer. Journal of Machine Learning Research, 21(140):1–67.
  12. Jialong Tang, Ziyao Lu, Jinsong Su, Yubin Ge, Linfeng Song, Le Sun, and Jiebo Luo. 2019. Progressive self-supervised attention learning for aspect-level sentiment analysis. In Proc. of ACL. find
  13. Bing Liu. 2012. Sentiment analysis and opinion mining. Synthesis lectures on human language technologies, 5(1):1–167. find
  14. David Wadden, Ulme Wennberg, Yi Luan, and Hannaneh Hajishirzi. 2019. Entity, relation, and event extraction with contextualized span representations. In Proc. of EMNLP.
  15. Yi Luan, Dave Wadden, Luheng He, Amy Shah, Mari Ostendorf, and Hannaneh Hajishirzi. 2019. A general framework for information extraction using dynamic span graphs. In Proc. of NAACL.
  16. Hai Wan, Yufei Yang, Jianfeng Du, Yanan Liu, Kunxun Qi, and Jeff Z. Pan. 2020. Target-aspect-sentiment joint detection for aspect-based sentiment analysis. In Proc. of AAAI. find
  17. Dehong Ma, Sujian Li, and Houfeng Wang. 2018. Joint learning for targeted sentiment analysis. In Proc. of EMNLP. find
  18. Alex Wang, Amanpreet Singh, Julian Michael, Felix Hill, Omer Levy, and Samuel Bowman. 2019. GLUE: A multi-task benchmark and analysis platform for natural language understanding. In Proc. of ICLR.
  19. Dehong Ma, Sujian Li, Fangzhao Wu, Xing Xie, and Houfeng Wang. 2019. Exploring sequence-to-sequence learning in aspect term extraction. In Proc. of ACL. find
  20. Haiyun Peng, Yukun Ma, Yang Li, and Erik Cambria. 2018. Learning multi-grained aspect target sequence for chinese sentiment analysis. Knowledge-Based Systems, 148:167–176. find
  21. Haiyun Peng, Lu Xu, Lidong Bing, Fei Huang, Wei Lu, and Luo Si. 2019. Knowing what, how and why: A near complete solution for aspect-based sentiment analysis. In Proc. of AAAI. find
  22. Jeffrey Pennington, Richard Socher, and Christopher D. Manning. 2014. Glove: Global vectors for word representation. In Proc. of EMNLP.
  23. Maria Pontiki, Dimitris Galanis, Haris Papageorgiou, Ion Androutsopoulos, Suresh Manandhar, Mohammed AL-Smadi, Mahmoud Al-Ayyoub, Yanyan Zhao, Bing Qin, Orphée De Clercq, et al. 2016. Semeval-2016 task 5: Aspect based sentiment analysis. In Proc. of SemEval.
  24. Bailin Wang and Wei Lu. 2018. Learning latent opinions for aspect-level sentiment classification. In Proc. of AAAI. find
  25. Wenya Wang, Sinno Jialin Pan, Daniel Dahlmeier, and Xiaokui Xiao. 2017. Coupled multi-layer attentions for co-extraction of aspect and opinion terms. In Proc. of AAAI. find
  26. Zhen Wu, Chengcan Ying, Fei Zhao, Zhifang Fan, Xinyu Dai, and Rui Xia. 2020. Grid tagging scheme for aspect-oriented fine-grained opinion extraction. In Findings of EMNLP. find
  27. Lu Xu, Lidong Bing, Wei Lu, and Fei Huang. 2020a. Aspect sentiment classification with aspect-specific opinion spans. In Proc. of EMNLP. find
  28. Lu Xu, Hao Li, W. Lu, and Lidong Bing. 2020b. Position-aware tagging for aspect sentiment triplet extraction. In Proc. of EMNLP.
  29. Bishan Yang and Claire Cardie. 2012. Extracting opinion expressions with semi-Markov conditional random fields. In Proc. of EMNLP. find
  30. Maria Pontiki, Dimitris Galanis, Haris Papageorgiou, Suresh Manandhar, and Ion Androutsopoulos. 2015. SemEval-2015 task 12: Aspect based sentiment analysis. In Proc. of SemEval.
  31. Bishan Yang and Claire Cardie. 2013. Joint inference for fine-grained opinion extraction. In Proc. of ACL.
  32. Maria Pontiki, Dimitris Galanis, John Pavlopoulos, Harris Papageorgiou, Ion Androutsopoulos, and Suresh Manandhar. 2014. Semeval-2014 task 4: Aspect based sentiment analysis. In Proc. of SemEval.
  33. Min Yang, Wenting Tu, Jingxuan Wang, Fei Xu, and Xiaojun Chen. 2017. Attention based lstm for target dependent sentiment classification. In Proc. of AAAI. find
  34. Guang Qiu, Bing Liu, Jiajun Bu, and Chun Chen. 2011. Opinion word expansion and target extraction through double propagation. Computational Linguistics, 37(1):9–27.
  35. Zhilin Yang, Zihang Dai, Yiming Yang, Jaime Carbonell, Ruslan Salakhutdinov, and Quoc V. Le. 2020. Xlnet: Generalized autoregressive pretraining for language understanding. In Proc. of NeurIPS.
  36. Yichun Yin, Furu Wei, Li Dong, Kaimeng Xu, Ming Zhang, and Ming Zhou. 2016. Unsupervised word and dependency path embeddings for aspect term extraction. In Proc. of IJCAI. find
  37. Chen Zhang, Qiuchi Li, Dawei Song, and Benyou Wang. 2020. A multi-task learning framework for opinion triplet extraction. In Findings of EMNLP. find
  38. Meishan Zhang, Yue Zhang, and Duy-Tin Vo. 2016. Gated neural networks for targeted sentiment analysis. In Proc. of AAAI. find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/xu-2021-learning-span-level-interactions-aspect/