Elamur Research
☰
home / papers / 2021 / Aspect Sentiment Quad Prediction as Paraphrase Generation

Aspect Sentiment Quad Prediction as Paraphrase Generation

Written for the Research library (claude-opus-5-5) from the paper text. Original — aclanthology.org/2021.emnlp-main.726.pdf

Summary

Aspect-based sentiment analysis (ABSA) is built around four elements: the category an aspect falls under, the term naming the aspect itself, the words expressing an opinion about it, and the polarity of the sentiment conveyed. Earlier tasks such as ASTE or TASD recover only part of this structure. The authors introduce Aspect Sentiment Quad Prediction (ASQP), which requires all four elements to be output together as quads, and release two datasets with quad-level annotations.

Pipeline systems accumulate errors across stages and treat labels as bare class indices. The authors instead propose PARAPHRASE modeling. Each quad is turned into a short templated sentence, roughly "category is great/bad because term is opinion". A pretrained T5 encoder-decoder is fine-tuned to generate these sentences from the input review, and the quads are then parsed back out of the generated text. This end-to-end setup lets the model exploit both the meaning of the labels and the knowledge stored in the pretrained model.

The same architecture surpasses earlier state-of-the-art systems on ASQP, ASTE and TASD. It also enables knowledge transfer between related tasks, which matters most when quad annotations are scarce.

Key points

  • Data: The new ASQP data covers restaurant reviews (Rest15, Rest16 splits). Code and annotations are public at github.com/IsakZhang/ABSA-QUAD.
  • TASD results (F1): PARAPHRASE scores 63.06 on Rest15 and 71.97 on Rest16. The generative baseline GAS scores 60.63 and 68.31, and MEJD scores 57.76 and 67.66.
  • Error analysis: On 100 sampled development sentences, opinion-term span boundaries were the most frequent mistake. Other errors included confusing similar categories (food quality vs. food style options) and confusing positive with neutral polarity. Outputs falling outside the allowed label vocabulary were rare.
  • Cross-task transfer: Adding 500 ASTE and TASD examples and training on just 5% of the ASQP data gave results comparable to training on 50% of the ASQP data without transfer. Even 100 transfer examples clearly helped in low-resource settings.

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: Machine translation · Natural language processing

Cited by

113 papers of our library cite it (of the 140673 whose bibliographies we read); 1 of them have a page here, 112 not yet.

Cite

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2021.emnlp-main.726

@inproceedings{zhang2021aspect,
  title = {{Aspect Sentiment Quad Prediction as Paraphrase Generation}},
  author = {Zhang, Wenxuan and Deng, Yang and Li, Xin and Yuan, Yifei and Bing, Lidong and Lam, Wai},
  booktitle = {Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing},
  year = {2021},
  doi = {10.18653/v1/2021.emnlp-main.726},
  url = {https://doi.org/10.18653/v1/2021.emnlp-main.726}
}

References

42 references, 2 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Ben Athiwaratkun, Cicero Nogueira dos Santos, Jason Krone, and Bing Xiang. 2020. Augmented natural language for generative sequence labeling. In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP), pages 375–385, Online. Association for Computational Linguistics. find
  2. Rahul Bhagat and Eduard H. Hovy. 2013. What is a paraphrase? Comput. Linguistics, 39(3):463–472. find
  3. Caroline Brun and Vassilina Nikoulina. 2018. Aspect based sentiment analysis into the wild. In Proceedings of the 9th Workshop on Computational Approaches to Subjectivity, Sentiment and Social Media Analysis, WASSA@EMNLP 2018, pages 116– 122. find
  4. Jiahao Bu, Lei Ren, Shuang Zheng, Yang Yang, Jingang Wang, Fuzheng Zhang, and Wei Wu. 2021. ASAP: A chinese review dataset towards aspect category sentiment analysis and rating prediction. CoRR, abs/2103.06605. ↗
  5. Hongjie Cai, Yaofeng Tu, Xiangsheng Zhou, Jianfei Yu, and Rui Xia. 2020. Aspect-category based sentiment analysis with hierarchical graph convolutional network. In Proceedings of the 28th International Conference on Computational Linguistics, COLING 2020, pages 833–843. find
  6. Shaowei Chen, Jie Liu, Yu Wang, Wenzheng Zhang, and Ziming Chi. 2020. Synchronous double-channel recurrent network for aspect-opinion pair extraction. In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, ACL 2020, pages 6515–6524. find
  7. We introduce a new ABSA task, namely aspect Shaowei Chen, Yu Wang, Jie Liu, and Yuelin Wang. sentiment quad prediction (ASQP) in this paper, 2021. Bidirectional machine reading comprehen-aiming to provide a more comprehensive aspect-sion for aspect sentiment triplet extraction. In AAAI. level sentiment picture. We propose a novel P ARA -Zhuang Chen and Tieyun Qian. 2020. Relation-aware PHRASE modeling paradigm that tackles the orig-collaborative learning for unified aspect-based sen-inal quad prediction as a paraphrase generation timent analysis. In Proceedings of the 58th Annual problem. Experiments on two datasets show its Meeting of the Association for Computational Linguistics, ACL 2020, pages 3685–3694. superiority compared with previous state-of-the-art models. We also demonstrate that the proposed Jacob Devlin, Ming-Wei Chang, Kenton Lee, and method provides a unified framework that can be Kristina Toutanova. 2019. BERT: Pre-training of easily adapted to handle other ABSA tasks as well. deep bidirectional transformers for language understanding. In NAACL, pages 4171–4186. Extensive analysis are conducted to understand the 9217 find
  8. Ruidan He, Wee Sun Lee, Hwee Tou Ng, and Daniel Dahlmeier. 2019. An interactive multi-task learning network for end-to-end aspect-based sentiment analysis. In ACL19, pages 504–515. find
  9. Mengting Hu, Shiwan Zhao, Li Zhang, Keke Cai, Zhong Su, Renhong Cheng, and Xiaowei Shen. 2019a. CAN: constrained attention networks for multi-aspect sentiment analysis. In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing, EMNLP-IJCNLP 2019, pages 4600–4609. find
  10. Minghao Hu, Yuxing Peng, Zhen Huang, Dongsheng Li, and Yiwei Lv. 2019b. Open-domain targeted sentiment analysis via span-based extraction and classification. In Proceedings of the 57th Conference of the Association for Computational Linguistics, ACL 2019, pages 537–546. find
  11. Binxuan Huang and Kathleen M. Carley. 2018. Parameterized convolutional neural networks for aspect level sentiment classification. In Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing, pages 1091–1096. find
  12. Lianzhe Huang, Peiyi Wang, Sujian Li, Tianyu Liu, Xiaodong Zhang, Zhicong Cheng, Dawei Yin, and Houfeng Wang. 2021. First target and opinion then polarity: Enhancing target-opinion correlation for aspect sentiment triplet extraction. CoRR, abs/2102.08549. ↗
  13. Xin Li, Lidong Bing, Piji Li, and Wai Lam. 2019a. A unified model for opinion target extraction and target sentiment prediction. In AAAI, pages 6714–6721. find
  14. Xin Li, Lidong Bing, Wenxuan Zhang, and Wai Lam. 2019b. Exploiting BERT for end-to-end aspect-based sentiment analysis. In Proceedings of the 5th Workshop on Noisy User-generated Text, W-NUT@EMNLP 2019, pages 34–41. find
  15. Bing Liu. 2012. Sentiment Analysis and Opinion Mining. Synthesis Lectures on Human Language Technologies. find
  16. Pengfei Liu, Shafiq R. Joty, and Helen M. Meng. 2015. Fine-grained opinion mining with recurrent neural networks and word embeddings. In Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing, EMNLP 2015, pages 1433–1443. find
  17. Xiao Liu, Yanan Zheng, Zhengxiao Du, Ming Ding, Yujie Qian, Zhilin Yang, and Jie Tang. 2021. GPT understands, too. CoRR, abs/2103.10385. ↗ our card
  18. Huaishao Luo, Tianrui Li, Bing Liu, and Junbo Zhang. 2019. DOER: dual cross-shared RNN for aspect term-polarity co-extraction. In ACL, pages 591– 601. find
  19. Dehong Ma, Sujian Li, Fangzhao Wu, Xing Xie, and Houfeng Wang. 2019. Exploring sequence-to-sequence learning in aspect term extraction. In Proceedings of the 57th Conference of the Association for Computational Linguistics, ACL 2019, pages 3538–3547. find
  20. Yue Mao, Yi Shen, Chao Yu, and Longjun Cai. 2021. A joint training dual-mrc framework for aspect based sentiment analysis. CoRR, abs/2101.00816. ↗
  21. Giovanni Paolini, Ben Athiwaratkun, Jason Krone, Jie Ma, Alessandro Achille, RISHITA ANUBHAI, Cicero Nogueira dos Santos, Bing Xiang, and Stefano Soatto. 2021. Structured prediction as translation between augmented natural languages. In International Conference on Learning Representations. find
  22. Haiyun Peng, Lu Xu, Lidong Bing, Fei Huang, Wei Lu, and Luo Si. 2020. Knowing what, how and why: A near complete solution for aspect-based sentiment analysis. In The Thirty-Fourth AAAI Conference on Artificial Intelligence, AAAI 2020, pages 8600– 8607. find
  23. Maria Pontiki, Dimitris Galanis, Haris Papageorgiou, Ion Androutsopoulos, Suresh Manandhar, Mohammad Al-Smadi, Mahmoud Al-Ayyoub, Yanyan Zhao, Bing Qin, Orphée De Clercq, Véronique Hoste, Marianna Apidianaki, Xavier Tannier, Natalia V. Loukachevitch, Evgeniy V. Kotelnikov, Núria Bel, Salud María Jiménez Zafra, and Gülsen Eryigit. 2016. Semeval-2016 task 5: Aspect based sentiment analysis. In Proceedings of the 10th International Workshop on Semantic Evaluation, SemEval@NAACL-HLT 2016, pages 19–30. find
  24. Maria Pontiki, Dimitris Galanis, Haris Papageorgiou, Suresh Manandhar, and Ion Androutsopoulos. 2015. Semeval-2015 task 12: Aspect based sentiment analysis. In SemEval@NAACL-HLT, pages 486–495. find
  25. Maria Pontiki, Dimitris Galanis, John Pavlopoulos, Harris Papageorgiou, Ion Androutsopoulos, and Suresh Manandhar. 2014. Semeval-2014 task 4: Aspect based sentiment analysis. In SemEval@COLING 2014, pages 27–35. find
  26. 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. J. Mach. Learn. Res., 21:140:1–140:67. find
  27. Sebastian Ruder, Parsa Ghaffari, and John G. Breslin. 2016. A hierarchical model of reviews for aspect-based sentiment analysis. In Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing, EMNLP 2016, pages 999– 1005. find
  28. Chi Sun, Luyao Huang, and Xipeng Qiu. 2019. Utilizing BERT for aspect-based sentiment analysis via constructing auxiliary sentence. In Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, NAACL-HLT 2019, pages 380–385. find our card
  29. Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N. Gomez, Lukasz Kaiser, and Illia Polosukhin. 2017. Attention is all you need. In Advances in Neural Information Processing Systems 30: Annual Conference on Neural Information Processing Systems 2017, pages 5998– 6008. find
  30. 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 The Thirty-Fourth AAAI Conference on Artificial Intelligence, AAAI 2020, pages 9122–9129. find
  31. Wenya Wang, Sinno Jialin Pan, Daniel Dahlmeier, and Xiaokui Xiao. 2017. Coupled multi-layer attentions for co-extraction of aspect and opinion terms. In Proceedings of the Thirty-First AAAI Conference on Artificial Intelligence, pages 3316–3322. find
  32. Yequan Wang, Minlie Huang, Xiaoyan Zhu, and Li Zhao. 2016. Attention-based LSTM for aspect-level sentiment classification. In Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing, EMNLP 2016, pages 606– 615. find
  33. Mi Zhang and Tieyun Qian. 2020. Convolution over hierarchical syntactic and lexical graphs for aspect level sentiment analysis. In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing, EMNLP 2020, pages 3540–3549. find
  34. Wenxuan Zhang, Xin Li, Yang Deng, Lidong Bing, and Wai Lam. 2021. Towards generative aspect-based sentiment analysis. In Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing, ACL/IJCNLP 2021, pages 504–510. find
  35. He Zhao, Longtao Huang, Rong Zhang, Quan Lu, and Hui Xue. 2020. Spanmlt: A span-based multi-task learning framework for pair-wise aspect and opinion terms extraction. In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, ACL 2020, pages 3239–3248. find
  36. Xinjie Zhou, Xiaojun Wan, and Jianguo Xiao. 2015. Representation learning for aspect category detection in online reviews. In Proceedings of the Twenty-Ninth AAAI Conference on Artificial Intelligence, pages 417–424. find
  37. Chao Wu, Qingyu Xiong, Hualing Yi, Yang Yu, Qiwu Zhu, Min Gao, and Jie Chen. 2021. Multiple-element joint detection for aspect-based sentiment analysis. Knowledge-Based Systems, page 107073. find
  38. 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 the Association for Computational Linguistics: EMNLP 2020. find
  39. Hu Xu, Bing Liu, Lei Shu, and Philip S. Yu. 2018. Double embeddings and cnn-based sequence labeling for aspect extraction. In Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics, ACL 2018, pages 592–598. find
  40. Lu Xu, Hao Li, Wei Lu, and Lidong Bing. 2020. Position-aware tagging for aspect sentiment triplet extraction. In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing, EMNLP 2020, pages 2339–2349. find
  41. Hang Yan, Junqi Dai, Tuo Ji, Xipeng Qiu, and Zheng Zhang. 2021. A unified generative framework for aspect-based sentiment analysis. In Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing, ACL/IJCNLP 2021, pages 2416–2429. find
  42. 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 Proceedings of the Twenty-Fifth International Joint Conference on Artificial Intelligence, IJCAI 2016, pages 2979–2985. find