Elamur Research⌕
☰
home / papers / 2018 / Aspect Based Sentiment Analysis with Gated Convolutional Networks

Aspect Based Sentiment Analysis with Gated Convolutional Networks

Paper · 2018 · Wei Xue and Tao Li

Summary

The paper addresses aspect based sentiment analysis (ABSA), which assigns a polarity to a specific aspect rather than to a whole review. The authors split ABSA into two subtasks. Aspect-category sentiment analysis (ACSA) deals with predefined categories such as service. Aspect-term sentiment analysis (ATSA) deals with target phrases that appear in the sentence. Most earlier systems combine LSTMs with attention, which makes them slow to train. As an alternative, the authors propose GCAE (Gated Convolutional network with Aspect Embedding). In GCAE, two parallel convolutional layers over word embeddings feed new Gated Tanh-ReLU Units (GTRU). The ReLU branch receives the aspect embedding and scales the sentiment features coming from the tanh branch. Max pooling and softmax then produce the final label. For ATSA, a small extra CNN encodes multi-word aspect terms. On SemEval restaurant and laptop data, GCAE matches or beats LSTM/attention models, with the largest gains on "hard" sentences containing several aspects with different sentiments. It also trains far faster. This shows that attention-free, parallelizable architectures are a viable choice for fine-grained sentiment.

Key points

  • Data: SemEval 2014 Task 4 (restaurants, laptops), plus Restaurant-Large, a merge of SemEval restaurant reviews from three years. Evaluation also uses hard test subsets of sentences with multiple aspects and mixed sentiment.
  • ACSA results: On Restaurant-Large, GCAE reaches 85.92 (test) and 70.75 (hard test), compared with 83.91 and 66.32 for ATAE-LSTM. On Restaurant 2014 hard test, GCAE scores 50.55 versus 45.62. However, the aspect-agnostic GCN scores slightly higher on the plain test set (79.67 vs 79.35).
  • ATSA results: GCAE scores 77.28/56.73 on restaurants and 69.14/47.06 on laptops, beating RAM, IAN, TD-LSTM and ATAE-LSTM. An SVM with large sentiment lexicons (NRC-Canada) remains stronger on the plain test sets (80.16 and 70.49).
  • Training speed: GCAE converges in about 3.33 s on ATSA, versus 82.87 s for IAN, 64.16 s for RAM and 25.28 s for ATAE-LSTM, all measured on a 1080 Ti GPU.
  • Gating ablation and code: GTRU outperforms the GLU and GTU gates, most clearly on Restaurant-Large hard test (70.75 vs about 60). The code and data are released at github.com/wxue004cs/GCAE.

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

🎓Florida International UniversityMiami · United States10 works · 1,508 citations

 Fields, topics and concepts

📚 Shelves 4The field of science this work belongs to (OpenAlex fields).Computer Science · 18,614 worksArtificial Intelligence · 17,965 worksArts and Humanities · 17,450 worksLanguage and Linguistics · 17,296 works

🧭 Topics 1Research subjects and languages the work is about.Sentiment & emotion · 1,035 works

Cite

Cite as Wei Xue and Tao Li. 2018. Aspect Based Sentiment Analysis with Gated Convolutional Networks. arXiv:1805.07043.

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.48550/arxiv.1805.07043 · arXiv 1805.07043

@misc{xue2018aspect,
  title = {{Aspect Based Sentiment Analysis with Gated Convolutional Networks}},
  author = {Wei Xue and Tao Li},
  year = {2018},
  doi = {10.48550/arxiv.1805.07043},
  eprint = {1805.07043},
  archivePrefix = {arXiv},
  url = {https://doi.org/10.48550/arxiv.1805.07043}
}

References

36 references, 18 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. 2014. Neural Machine Translation by Jointly Learning to Align and Translate. In ICLR, pages CoRR abs–1409.0473.
  2. Peng Chen, Zhongqian Sun, Lidong Bing, and Wei Yang. 2017. Recurrent Attention Network on Memory for Aspect Sentiment Analysis. In EMNLP, pages 463–472.
  3. Junyoung Chung, Caglar Gulcehre, Kyunghyun Cho, and Yoshua Bengio. 2014. Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling. In NIPS. find
  4. Ronan Collobert, Jason Weston, Léon Bottou, Michael Karlen, Koray Kavukcuoglu, and Pavel P Kuksa. 2011. Natural Language Processing (Almost) from Scratch. Journal of Machine Learning Research, 12:2493–2537.
  5. Alexis Conneau, Holger Schwenk, Loı̈c Barrault, and Yann LeCun. 2016. Very Deep Convolutional Networks for Text Classification. In EACL, pages 1107–1116.
  6. Yann N Dauphin, Angela Fan, Michael Auli, and David Grangier. 2017. Language Modeling with Gated Convolutional Networks. In ICML, pages 933–941.
  7. 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 ACL, pages 49–54.
  8. John C Duchi, Elad Hazan, and Yoram Singer. 2011. Adaptive Subgradient Methods for Online Learning and Stochastic Optimization. Journal of Machine Learning Research, pages 2121–2159. find
  9. Jonas Gehring, Michael Auli, David Grangier, Denis Yarats, and Yann N Dauphin. 2017. Convolutional Sequence to Sequence Learning. In ICML, pages 1243–1252.
  10. Sepp Hochreiter and Jürgen Schmidhuber. 1997. Long Short-Term Memory. Neural computation, 9(8):1735–1780. find
  11. Nal Kalchbrenner, Lasse Espeholt, Karen Simonyan, Aäron van den Oord, Alex Graves, and Koray Kavukcuoglu. 2016. Neural Machine Translation in Linear Time . CoRR, abs/1610.10099. find
  12. Nal Kalchbrenner, Edward Grefenstette, and Phil Blunsom. 2014. A convolutional neural network for modelling sentences. In ACL, pages 655–665.
  13. Yoon Kim. 2014. Convolutional Neural Networks for Sentence Classification. In EMNLP, pages 1746– 1751.
  14. Svetlana Kiritchenko, Xiaodan Zhu, Colin Cherry, and Saif M. Mohammad. 2014. NRC-Canada-2014: Detecting aspects and sentiment in customer reviews. In SemEval@COLING, pages 437–442, Stroudsburg, PA, USA. Association for Computational Linguistics.
  15. Siwei Lai, Liheng Xu, Kang Liu, and Jun Zhao. 2015. Recurrent Convolutional Neural Networks for Text Classification. AAAI, pages 2267–2273. find
  16. Himabindu Lakkaraju, Richard Socher, and Chris Manning. 2014. Aspect specific sentiment analysis using hierarchical deep learning. In NIPS Workshop on Deep Learning and Representation Learning. find
  17. Hoa T Le, Christophe Cerisara, and Alexandre Denis. 2017. Do Convolutional Networks need to be Deep for Text Classification ? CoRR, abs/1707.04108. — [arXiv:1707.04108](https://arxiv.org/abs/1707.04108) find
  18. Bing Liu and Lei Zhang. 2012. A Survey of Opinion Mining and Sentiment Analysis. Mining Text Data, (Chapter 13):415–463. find
  19. Dehong Ma, Sujian Li, Xiaodong Zhang, and Houfeng Wang. 2017. Interactive Attention Networks for Aspect-Level Sentiment Classification. In IJCAI, pages 4068–4074. International Joint Conferences on Artificial Intelligence Organization. find
  20. Aäron van den Oord, Nal Kalchbrenner, Lasse Espeholt, Koray Kavukcuoglu, Oriol Vinyals, and Alex Graves. 2016. Conditional Image Generation with PixelCNN Decoders. In NIPS, pages 4790–4798. find
  21. Bo Pang and Lillian Lee. 2008. Opinion mining and sentiment analysis. Foundations and Trends R in Information Retrieva, 2:1–135. find
  22. Jeffrey Pennington, Richard Socher, and Christopher D Manning. 2014. Glove: Global Vectors for Word Representation. In EMNLP, pages 1532–1543.
  23. Maria Pontiki, Dimitrios Galanis, John Pavlopoulos, Haris Papageorgiou, Ion Androutsopoulos, and Suresh Manandhar. 2014. Semeval-2014 task 4: Aspect based sentiment analysis. In SemEval@COLING, pages 27–35, Stroudsburg, PA, USA. Association for Computational Linguistics.
  24. Sebastian Ruder, Parsa Ghaffari, and John G Breslin. 2016a. A Hierarchical Model of Reviews for Aspect-based Sentiment Analysis. In EMNLP, pages 999–1005. find
  25. Sebastian Ruder, Parsa Ghaffari, and John G Breslin. 2016b. INSIGHT-1 at SemEval-2016 Task 5 - Deep Learning for Multilingual Aspect-based Sentiment Analysis. In SemEval@NAACL-HLT, pages 330– 336. find
  26. Richard Socher, Alex Perelygin, Jean Y Wu, and Jason Chuang. 2013. Recursive deep models for semantic compositionality over a sentiment treebank. In EMNLP, pages 1631–1642.
  27. Kai Sheng Tai, Richard Socher, and Christopher D Manning. 2015. Improved Semantic Representations From Tree-Structured Long Short-Term Memory Networks. In ACL, pages 1556–1566. find
  28. Duyu Tang, Bing Qin, Xiaocheng Feng, and Ting Liu. 2016a. Effective LSTMs for Target-Dependent Sentiment Classification. In COLING, pages 3298– 3307.
  29. Duyu Tang, Bing Qin, and Ting Liu. 2015. Document Modeling with Gated Recurrent Neural Network for Sentiment Classification. In EMNLP, pages 1422– 1432.
  30. Duyu Tang, Bing Qin, and Ting Liu. 2016b. Aspect Level Sentiment Classification with Deep Memory Network. In EMNLP, pages 214–224.
  31. Wenya Wang, Sinno Jialin Pan, Daniel Dahlmeier, and Xiaokui Xiao. 2016a. Recursive Neural Conditional Random Fields for Aspect-based Sentiment Analysis. In EMNLP, pages 616–626.
  32. Yequan Wang, Minlie Huang, Xiaoyan Zhu, and Li Zhao. 2016b. Attention-based LSTM for Aspect-level Sentiment Classification. In EMNLP, pages 606–615.
  33. Jason Weston, Sumit Chopra, and Antoine Bordes. 2014. Memory Networks. In ICLR, pages CoRR abs–1410.3916. find
  34. Jiacheng Xu, Danlu Chen, Xipeng Qiu, and Xuanjing Huang. 2016. Cached Long Short-Term Memory Neural Networks for Document-Level Sentiment Classification. In EMNLP, pages 1660–1669. find
  35. Wei Xue, Wubai Zhou, Tao Li, and Qing Wang. 2017. Mtna: A neural multi-task model for aspect category classification and aspect term extraction on restaurant reviews. In Proceedings of the Eighth International Joint Conference on Natural Language Processing (Volume 2: Short Papers), volume 2, pages… find
  36. Meishan Zhang, Yue Zhang, and Duy-Tin Vo. 2016. Gated Neural Networks for Targeted Sentiment Analysis. In AAAI, pages 3087–3093. find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/xue-2018-aspect-based-sentiment-analysis-gated/