ABSA 和 ASTE 任务简介 情感分析作为自然语言理解里最重要也是最有挑战的主要任务之一,有很大的研究空间和广阔的应用价值。细粒度的情感分析(i.e. Aspect-Based Sentiment Analysis(ABSA))的相关研究已历经多年,有大量的相关论文发表。 先给一个 ABSA 的例子,对于一句餐馆评论:“Waiters are very friendly and the pasta is simply average.”,提到了两个评论目标:“waiter”和“pasta”;用来评价他们的词分别是:“friendly”和“average”;这句话评论的分别是餐馆的“service”和“food”方面。 以上便是 ABSA 任务中处理的三类对象:aspect、opinion、aspect category。也就是下图中的最上层。从目标识别角度,针对 aspect term 和 opinion term,存在抽取问题;针对 aspect category,存在分类问题(假设预定义 aspect categories)。 从情感分析角度,对 aspect term 和 aspect category 存在情感分类问题。这些原子任务如下图中间层所示。注意,一句评论里可能没有显示提及 aspect term,但同样可以存在 aspect category,比如“I was treated rudely.”讲的是“service”。
[1] A Unified Model for Opinion Target Extraction and Target Sentiment Prediction. Xin Li, Lidong Bing, Piji Li, Wai Lam. The Thirty-Third AAAI Conference on Artificial Intelligence (AAAI'19), 2019.
[3] Knowing What, How and Why: A Near Complete Solution for Aspect-based Sentiment Analysis. Haiyun Peng, Lu Xu, Lidong Bing, Wei Lu, Fei Huang, Luo Si. The Thirty-Fourth AAAI Conference on Artificial Intelligence (AAAI'20), 2020.
[4] Position-Aware Tagging for Aspect Sentiment Triplet Extraction. Lu Xu, Hao Li, Wei Lu, Lidong Bing. The Conference on Empirical Methods in Natural Language Processing (EMNLP'20), 2020.
[5] Learning Span-Level Interactions for Aspect Sentiment Triplet Extraction. Lu Xu, Yew Ken Chia, Lidong Bing. The Joint Conference of the 59th Annual Meeting of the Association for Computational Linguistics (ACL'21), 2021.
ASTE任务的提出 本小节工作来自论文:
Knowing What, How and Why: A Near Complete Solution for Aspect-based Sentiment Analysis, In AAAI 2020.
[1] 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.
[2] He, R.; Lee, W. S.; Ng, H. T.; and Dahlmeier, D. 2019. An interactive multi-task learning network for end-to-end aspect-based sentiment analysis. In ACL.
[3] Wang, W.; Pan, S. J.; Dahlmeier, D.; and Xiao, X. 2017. Coupled multi-layer attentions for co-extraction of aspect and opinion terms. In AAAI, 3316–3322.
[4] Dai, H., and Song, Y. 2019. Neural aspect and opinion term extraction with mined rules as weak supervision. In ACL, 5268–5277.
[5] Hu, M.; Zhao, S.; Zhang, L.; Cai, K.; Su, Z.; Cheng, R.; and Shen, X. 2018. Can: Constrained attention networks for multi-aspect sentiment analysis. arXiv preprint arXiv:1812.10735.
[6] Fan, Z.; Wu, Z.; Dai, X.; Huang, S.; and Chen, J. 2019. Target-oriented opinion words extraction with target-fused neural sequence labeling. In NAACL-HLT, 2509–2518.
[7] Zhang, X., and Goldwasser, D. 2019. Sentiment tagging with partial labels using modular architectures. arXiv preprint arXiv:1906.00534.
JET: End-to-End ASTE
本小节工作来自论文:
Position-Aware Tagging for Aspect Sentiment Triplet Extraction, in EMNLP 2020
[1] Peng, H., Xu, L., Bing, L., Huang, F., Lu, W., & Si, L. (2020). Knowing What, How and Why: A Near Complete Solution for Aspect-Based Sentiment Analysis. Proceedings of the AAAI Conference on Artificial Intelligence, 34(05), 8600-8607.
[2] Wang, W.; Pan, S. J.; Dahlmeier, D.; and Xiao, X. 2017. Coupled multi-layer attentions for co-extraction of aspect and opinion terms. In AAAI, 3316–3322.
[3] Dai, H., and Song, Y. 2019. Neural aspect and opinion term extraction with mined rules as weak supervision. In ACL, 5268–5277.
[4] John Lafferty, Andrew McCallum, and Fernando CN Pereira. 2001. Conditional random fields: Probabilistic models for segmenting and labeling sequence data. In ICML.
[5] Sunita Sarawagi and William W Cohen. 2004. Semi– markov conditional random fields for information extraction. In NeurIPS.
Span-based ASTE
本小节工作来自论文:
Learning Span-Level Interactions for Aspect Sentiment Triplet Extraction, in ACL 2021
[1] Position-Aware Tagging for Aspect Sentiment Triplet Extraction. Lu Xu, Hao Li, Wei Lu, Lidong Bing. The Conference on Empirical Methods in Natural Language Processing (EMNLP'20), 2020.
[2] 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.
[3] Jeffrey Pennington, Richard Socher, and Christopher D. Manning. 2014. Glove: Global vectors for word representation. In Proc. of EMNLP.
[4] Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2019. BERT: Pre-training of deep bidirectional transformers for language understanding. In Proc. of NAACL.
[5] 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.