Elamur Research⌕
☰
home / papers / 2019 / Open-Domain Targeted Sentiment Analysis via Span-Based Extraction and Classification

Open-Domain Targeted Sentiment Analysis via Span-Based Extraction and Classification

Paper · 2019 · Minghao Hu, Yuxing Peng, Zhen Huang et al.

Summary

The paper tackles open-domain targeted sentiment analysis: finding the opinion targets in a sentence and assigning each one a sentiment polarity. Earlier systems treated this as sequence tagging with word-level labels. The authors argue that tagging has two weaknesses. Its search space is very large, and the words of one multi-word target can receive conflicting polarities.

They replace tagging with span-based labeling, in which each target is marked by its start position, end position and polarity. On top of this they build an extract-then-classify framework with a BERT backbone. A multi-target extractor scores start and end positions and uses a heuristic multi-span decoder to propose candidate targets. A polarity classifier then labels each candidate from its pooled span representation. They compare pipeline, joint and collapsed variants of this design.

On three benchmarks (LAPTOP, REST, TWITTER), the span-based models beat a CRF-based tagging baseline. The pipeline variant does best overall. This shows that boundary prediction, borrowed from extractive question answering, suits targeted sentiment better than tagging.

Key points

  • Decoding: Candidate spans need start plus end scores above a threshold, with a penalty for span length. Overlapping spans are then pruned by non-maximum suppression (overlap measured with word-level F1) until top-K targets remain.
  • Polarity classification: Span-level classification beats the tagging baseline by 9.97, 8.15 and 15.4 accuracy points. Accuracies are 81.39 vs 71.42 on LAPTOP, 89.95 vs 81.80 on REST, and 75.16 vs 59.76 on TWITTER. On LAPTOP it also exceeds MGAN (75.39) and TNet (76.54).
  • Target length: The tagger's accuracy falls sharply as targets get longer, while the span classifier stays stable. This confirms the sentiment-inconsistency problem of tagging.
  • Variants: The pipeline variant consistently outperforms the joint and collapsed variants. Most of the gain comes from the polarity classifier, and the extractor helps most on long sentences.
  • Code: Released at github.com/huminghao16/SpanABSA. The model uses BERT's 30,522-wordpiece vocabulary.

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

🎓University of MacauMacao · China10 works · 1,905 citations🎓National University of Defense TechnologyChangsha · China4 works · 918 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 Minghao Hu, Yuxing Peng, Zhen Huang, Dongsheng Li and Yiwei Lv. 2019. Open-Domain Targeted Sentiment Analysis via Span-Based Extraction and Classification. Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics.

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/p19-1051

@inproceedings{hu2019open,
  title = {{Open-Domain Targeted Sentiment Analysis via Span-Based Extraction and Classification}},
  author = {Minghao Hu, Yuxing Peng, Zhen Huang, Dongsheng Li and Yiwei Lv},
  booktitle = {Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics},
  year = {2019},
  doi = {10.18653/v1/p19-1051},
  url = {https://doi.org/10.18653/v1/p19-1051}
}

References

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

  1. Kenton Lee, Shimi Salant, Tom Kwiatkowski, Ankur Parikh, Dipanjan Das, and Jonathan Berant. 2016. Learning recurrent span representations for extractive question answering. arXiv preprint arXiv:1611.01436. — [arXiv:1611.01436](https://arxiv.org/abs/1611.01436) find
  2. Kenton Lee, Luheng He, Mike Lewis, and Luke Zettlemoyer. 2017. End-to-end neural coreference resolution. arXiv preprint arXiv:1707.07045.
  3. Dzmitry Bahdanau, Kyunghyun Cho, and Yoshua Bengio. 2014. Neural machine translation by jointly learning to align and translate. arXiv preprint arXiv:1409.0473.
  4. Xin Li, Lidong Bing, Wai Lam, and Bei Shi. 2018. Transformation networks for target-oriented sentiment classification. In Proceedings of ACL.
  5. Peng Chen, Zhongqian Sun, Lidong Bing, and Wei Yang. 2017. Recurrent attention network on memory for aspect sentiment analysis. In Proceedings of EMNLP.
  6. Xin Li, Lidong Bing, Piji Li, and Wai Lam. 2019. A unified model for opinion target extraction and target sentiment prediction. In Proceedings of AAAI. find
  7. Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2018. Bert: Pre-training of deep bidirectional transformers for language understanding. arXiv preprint arXiv:1810.04805.
  8. Xin Li and Wai Lam. 2017. Deep multi-task learning for aspect term extraction with memory interaction. In Proceedings of EMNLP. find
  9. Li Dong, Furu Wei, Chuanqi Tan, Duyu Tang, Ming Zhou, and Ke Xu. 2014. Adaptive recursive neural network for target-dependent twitter sentiment classification. In Proceedings of ACL.
  10. Chenghua Lin and Yulan He. 2009. Joint sentiment/topic model for sentiment analysis. In Proceedings of CIKM. find
  11. Feifan Fan, Yansong Feng, and Dongyan Zhao. 2018. Multi-grained attention network for aspect-level sentiment classification. In Proceedings of EMNLP.
  12. Bing Liu. 2012. Sentiment analysis and opinion mining. Synthesis lectures on human language technologies, 5(1):1–167. find
  13. Luheng He, Kenton Lee, Omer Levy, and Luke Zettlemoyer. 2018. Jointly predicting predicates and arguments in neural semantic role labeling. arXiv preprint arXiv:1805.04787.
  14. Pengfei Liu, Shafiq Joty, and Helen Meng. 2015. Fine-grained opinion mining with recurrent neural networks and word embeddings. In Proceedings of EMNLP.
  15. Ruidan He, Wee Sun Lee, Hwee Tou Ng, and Daniel Dahlmeier. 2017. An unsupervised neural attention model for aspect extraction. In Proceedings of ACL.
  16. Margaret Mitchell, Jacqui Aguilar, Theresa Wilson, and Benjamin Van Durme. 2013. Open domain targeted sentiment. In Proceedings of EMNLP.
  17. Sepp Hochreiter and Jürgen Schmidhuber. 1997. Long short-term memory. Neural computation. find
  18. Bo Pang, Lillian Lee, et al. 2008. Opinion mining and sentiment analysis. Foundations and Trends R in Information Retrieval, 2(1–2):1–135. find
  19. Minghao Hu, Yuxing Peng, Zhen Huang, Xipeng Qiu, Furu Wei, and Ming Zhou. 2018. Reinforced mnemonic reader for machine reading comprehension. In Proceedings of IJCAI. find
  20. Binxuan Huang and Kathleen Carley. 2018. Parameterized convolutional neural networks for aspect level sentiment classification. In Proceedings of EMNLP. find
  21. Maria Pontiki, Dimitris Galanis, Haris Papageorgiou, Ion Androutsopoulos, Suresh Manandhar, AL-Smadi Mohammad, Mahmoud Al-Ayyoub, Yanyan Zhao, Bing Qin, Orphée De Clercq, et al. 2016. Semeval-2016 task 5: Aspect based sentiment analysis. In Proceedings of SemEval-2016.
  22. Niklas Jakob and Iryna Gurevych. 2010. Extracting opinion targets in a single-and cross-domain setting with conditional random fields. In Proceedings of EMNLP.
  23. Maria Pontiki, Dimitris Galanis, Haris Papageorgiou, Suresh Manandhar, and Ion Androutsopoulos. 2015. Semeval-2015 task 12: Aspect based sentiment analysis. In Proceedings of SemEval 2015.
  24. Maria Pontiki, Dimitris Galanis, John Pavlopoulos, Harris Papageorgiou, Ion Androutsopoulos, and Suresh Manandhar. 2014. Semeval-2014 task 4: Aspect based sentiment analysis. In Proceedings of SemEval-2014.
  25. Soujanya Poria, Erik Cambria, and Alexander Gelbukh. 2016. Aspect extraction for opinion mining with a deep convolutional neural network. Knowledge-Based Systems, 108:42–49. find
  26. Pranav Rajpurkar, Jian Zhang, Konstantin Lopyrev, and Percy Liang. 2016. Squad: 100,000+ questions for machine comprehension of text. In Proceedings of EMNLP.
  27. Azriel Rosenfeld and Mark Thurston. 1971. Edge and curve detection for visual scene analysis. IEEE Transactions on computers, (5):562–569. find
  28. Minjoon Seo, Aniruddha Kembhavi, Ali Farhadi, and Hannaneh Hajishirzi. 2017. Bidirectional attention flow for machine comprehension. In Proceedings of ICLR.
  29. Lei Shu, Hu Xu, and Bing Liu. 2017. Lifelong learning crf for supervised aspect extraction. In Proceedings of the ACL. find
  30. Duyu Tang, Bing Qin, Xiaocheng Feng, and Ting Liu. 2016a. Effective lstms for target-dependent sentiment classification. In Proceedings of COLING.
  31. Duyu Tang, Bing Qin, and Ting Liu. 2016b. Aspect level sentiment classification with deep memory network. arXiv preprint arXiv:1605.08900.
  32. Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez, Łukasz Kaiser, and Illia Polosukhin. 2017. Attention is all you need. In Proceedings of NIPS.
  33. Shuohang Wang and Jing Jiang. 2017. Machine comprehension using match-lstm and answer pointer. In Proceedings of ICLR.
  34. Wenya Wang, Sinno Jialin Pan, Daniel Dahlmeier, and Xiaokui Xiao. 2016a. Recursive neural conditional random fields for aspect-based sentiment analysis. In Proceedings of EMNLP.
  35. Yequan Wang, Minlie Huang, Li Zhao, et al. 2016b. Attention-based lstm for aspect-level sentiment classification. In Proceedings of EMNLP.
  36. Hu Xu, Bing Liu, Lei Shu, and Philip S Yu. 2018. Double embeddings and cnn-based sequence labeling for aspect extraction. In Proceedings of ACL.
  37. Wei Xue and Tao Li. 2018. Aspect based sentiment analysis with gated convolutional networks. arXiv preprint arXiv:1805.07043. find
  38. Meishan Zhang, Yue Zhang, and Duy Tin Vo. 2015. Neural networks for open domain targeted sentiment. In Proceedings of EMNLP. find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/hu-2019-open-domain-targeted-sentiment-analysis/