Abstract
End-to-end aspect-based sentiment analysis aims to complete aspect terms extraction and aspect sentiment classification simultaneously. Most existing methods ignore the sematic connection between the two subtasks. In this paper, we solve the problem by inducing constituents from input sentences, and propose a novel model based on sentence constituent-aware attention mechanism for end-to-end aspect-based sentiment analysis. Our framework mainly involves three layers. The first layer gets word representations by the pre-trained language model. Followed by the proposed sentence constituent-aware attention layer to induce constituents from the input sentence. With the operation of inducing constituents, the words in the same constituent are constrained to attend to each other, making the aspect term pay more attention to its corresponding opinion. Finally, a simple linear classification layer is adopted to predict the unified tags. Experimental results demonstrate that the proposed model outperforms other baselines on four benchmark datasets.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
References
Bie Y, Yang Y (2021) A multitask multi-view neural network for end-to-end aspect-based sentiment analysis. Big Data Mining and Analytics 4(3):195–207
Chen P, Sun Z, Bing L, Yang W (2017) Recurrent attention network on memory for aspect sentiment analysis. In: EMNLP, pp 452–461
Fan F, Feng Y, Zhao D (2018) Multi-grained attention network for aspect-level sentiment classification. In: EMNLP, pp 3433–3442
He R, Lee WS, Ng HT, Dahlmeier D (2017) An Unsupervised Neural Attention Model for Aspect Extraction. In: ACL, pp 388–397
He R, Lee WS, Ng HT, Dahlmeier D (2018) Exploiting document knowledge for aspect-level sentiment classification. In: ACL, pp 579–585
He R, Lee WS, Ng HT, Dahlmeier D (2019) An interactive multi-task learning network for end-to-end aspect-based sentiment analysis. In: ACL, pp 504–515
Li X, Lam W (2017) Deep multi-task learning for aspect term extraction with memory interaction. In: EMNLP, pp 2886–2892
Li X, Bing L, Li P, Lam W, Yang Z (2018a) Aspect term extraction with history attention and selective transformation. In: IJCAI, pp 4194–4200
Li X, Bing L, Lam W, Shi B (2018b) Transformation networks for target-oriented sentiment classification. In: ACL, pp 946–956
Li X, Bing L, Li P, Lam W (2019a) A unified model for opinion target extraction and target sentiment prediction. In: AAAI, pp 6714–6721
Li X, Bing L, Zhang W, Lam W (2019b) Exploiting BERT for End-to-End Aspect-based Sentiment Analysis. In: EMNLP, pp 34–41
Li K, Chen C, Quan X, Ling Q, Song Y (2020) Conditional Augmentation for Aspect Term Extraction via Masked Sequence-to-Sequence Generation. In: ACL, pp 7056–7066
Liang Y, Meng F, Zhang J, Xu J, Chen Y, Zhou J (2021) A dependency syntactic knowledge augmented interactive architecture for end-to-end aspect-based sentiment analysis. Neurocomputing 454:291–302
Liu L, Shang J, Ren X, Frank FX, Gui H, Peng J, Han J (2018) Empower sequence labeling with task-aware neural language model. In: AAAI, pp 5253–5260
Luo H, Li T, Liu B, Wang B, Unger H (2018) Improving aspect term extraction with bidirectional dependency tree representation. arXiv:1805.07889
Luo H, Li T, Liu B, Zhang J (2019) DOER: Dual cross-shared RNN for aspect term-polarity co-extraction. In: ACL, pp 591–601
Ma D, Li S, Zhang X, Wang H (2017) Interactive attention networks for aspect-level sentiment classification. In: IJCAI, pp 4068–4074
Mitchell M, Aguilar J, Wilson T, Van Durme B (2013) Open domain targeted sentiment. In: EMNLP, pp 1643–1654
Rostami M, Berahmand K, Forouzandeh S (2020) A novel method of constrained feature selection by the measurement of pairwise constraints uncertainty. J Big Data 7:83
Rostami M, Berahmand K, Forouzandeh S (2020) A novel community detection based genetic algorithm for feature selection. CoRR abs/2008.03543
Rostami M, Berahmand K, Nasiri E, Forouzandeh S (2021) Review of swarm intelligence-based feature selection methods. Eng Appl Artif Intell 100:104210
Tang D, Qin B, Liu T (2016b) Aspect level sentiment classification with deep memory network. In: EMNLP, pp 214–224
Wang B, Lu W (2018) Learning latent opinions for aspect-level sentiment classification. In: AAAI, pp 5537–5544
Wang Y, Huang M, Zhao L et al (2016) Attention-based lstm for aspect-level sentiment classification. In: EMNLP, pp 606–615
Wang S, Mazumder S, Liu B, Zhou M, Chang Y (2018) Target-sensitive memory networks for aspect sentiment classification. In: ACL, pp 957–967
Wang X, Xu G, Zhang Z, Jin L, Sun X (2021) End-to-end aspect-based sentiment analysis with hierarchical multi-task learning. Neurocomputing 455:178–188
Xu H, Liu B, Shu L, Yu PS (2018) Double embeddings and cnn-based sequence labeling for aspect extraction. In: ACL, pp 592–598
Xu L, Bing L, Lu W, Huang F (2020) Aspect sentiment classification with aspect-specific opinion spans. In: EMNLP, pp 3561–3567
Xue W, Li T (2018) Aspect Based Sentiment Analysis with Gated Convolutional Networks. In: ACL, pp 2514–2523
Zhang M, Zhang Y, Vo DT (2015) Neural networks for open domain targeted sentiment. In: EMNLP, pp 612–621
Zheng L, Wei Y, Yu Z, Zhang X, Li X (2019) Exploiting coarse-to-fine task transfer for aspect-level sentiment classification. In: AAAI, vol 33, pp 4253–4260
Acknowledgments
This work was supported by National Natural Science Foundation of China (62162037) and General Projects of Basic Research in Yunnan Province (202001AT070047).
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Lu, T., Xiang, Y., Zhang, L. et al. Sentence constituent-aware attention mechanism for end-to-end aspect-based sentiment analysis. Multimed Tools Appl 81, 15333–15348 (2022). https://doi.org/10.1007/s11042-022-12487-x
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11042-022-12487-x