基于多任务学习注意力交互模型的方面级情感分析
Aspect-Level Sentiment Analysis Based on Multi-Task Learning Attention Interaction Model
DOI: 10.12677/CSA.2022.121002, PDF,    科研立项经费支持
作者: 叶洲铭, 王 瑛：广东工业大学，计算机学院，广东 广州
关键词: 方面级情感分析；多头注意力机制；多任务模型；Aspect-Level Sentiment Analysis； Multi-Head Attention Mechanism； Multi-Task Model
摘要: 基于方面级的情感分析，其现有算法大多是将方面词提取和情感极性分类分开独立处理，或是只执行其中一个任务，忽略两个子任务间潜在的关联；或是使用流水线方法分为两个阶段分别处理，导致系统误差传播以及复杂化。本文将两个任务统一定义为序列标注问题，先使用Bi-LSTM和CNN提取特征，其次使用基于多头注意力的特征交互学习机制对特征进一步筛选提取；并且使用一个基于注意力交互机制，获得任务间潜在关联特征。根据实验表明，所提出的算法在3个公开数据集上，在情感极性分类以及方面词提取上，比当前代表模型的准确率有所提高。
Abstract: Based on aspect sentiment analysis, most of its existing algorithms separate aspect word extraction and sentiment polarity classification separately, or perform only one of the tasks, ignoring the potential correlation between the two subtasks; or use pipeline method to analyze. They are processed separately for two stages, leading to system error propagation and complexity. This paper defines the two tasks as a sequence labeling problem. First, Bi-LSTM and CNN are used to extract features, and then the feature interactive learning mechanism based on multi-head attention is used to further filter and extract features; and an attention-based interaction mechanism is used to obtain the task potentially related features. According to experiments, the proposed algorithm is more accurate than the current representative model in terms of emotion polarity classification and aspect word extraction on three public data sets.
文章引用：叶洲铭, 王瑛, 王勇. 基于多任务学习注意力交互模型的方面级情感分析[J]. 计算机科学与应用, 2022, 12(1): 10-16. https://doi.org/10.12677/CSA.2022.121002

参考文献

[1] Yadav, A. and Vishwakarma, D.K. (2020) Sentiment Analysis Using Deep Learning Architectures: A Review. Artificial Intelligence Review, 53, 4335-4385. [Google Scholar] [CrossRef] 
[2] Zhou, J., Huang, J.X., Chen, Q., et al. (2019) Deep Learning for Aspect-Level Sentiment Classification: Survey, Vision and Challenges. IEEE Access, 7, 78454-78483. [Google Scholar] [CrossRef] 
[3] 郝亚洲, 郑庆华, 陈艳平, 闫彩霞. 面向网络舆情数据的异常行为识别[J]. 计算机研究与发展, 2016, 53(3): 611-620.
[4] 张宜浩, 朱小飞, 徐传运, 董世都. 基于用户评论的深度情感分析和多视图协同融合的混合推荐方法[J]. 计算机学报, 2019, 42(6): 1316-1333.
[5] Devlin, J., Chang, M.W., Lee, K., et al. (2019) Bert: Pre-Training of Deep Bidirectional Transformers for Language Understanding. Proceeding of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Minneapolis, 2-7 June 2019, 4171-4186.
[6] Song, Y., Wang, J., Tao, J., et al. (2019) Attentional Encoder Network for Targeted Sentiment Classification.
[7] Zeng, B., Yang, H., Xu, R., et al. (2019) LCF: A Local Context Focus Mechanism for Aspect-Based Sentiment Classification. Applied Sciences, 9, 3389-3389. [Google Scholar] [CrossRef] 
[8] Tang, D., Qin, B., Feng, X., et al. (2015) Effec-tive LSTMs for Target-Dependent Sentiment Classification. COLING 2016, 26th International Conference on Computa-tional Linguistics, Proceedings of the Conference: Technical Papers, Osaka, 11-17 December 2016, 3298-3307.
[9] 彭祝亮, 刘博文, 范程岸, 王杰, 肖明, 廖泽恩. 基于BLSTM与方面注意力模块的情感分类方法[J]. 计算机工程, 2020, 46(3): 60-65+72. [Google Scholar] [CrossRef] 
[10] Dong, L., Wei, F., Tan, C., et al. (2014) Adaptive Recursive Neural Network for Target-Dependent Twitter Sentiment Classification. Proceedings of the 52nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers), Baltimore, 23-25 June 2014, 49-54. [Google Scholar] [CrossRef] 
[11] Liu, P., Joty, S. and Meng, H. (2015) Fine-grained Opinion Mining with Recurrent Neural Networks and Word Embeddings. Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing, Lisbon, 17-21 September 2015, 1433-1443. [Google Scholar] [CrossRef] 
[12] Ye, H., Yan, Z., Luo, Z., et al. (2017) Dependency-Tree Based Con-volutional Neural Networks for Aspect Term Extraction. Pacific-Asia Conference on Knowledge Discovery and Data Mining, Jeju, 23-26 May 2017, 350-362. [Google Scholar] [CrossRef] 
[13] Wang, Y., Huang, M., Zhu, X., et al. (2016) Attention-Based LSTM for Aspect-Level Sentiment Classification. Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing, Austin, 1-5 November 2016, 606-615. [Google Scholar] [CrossRef] 
[14] Wang, F., Lan, M. and Wang, W. (2018) Towards a One-Stop Solution to Both Aspect Extraction and Sentiment Analysis Tasks with Neural Multi-task Learning. 2018 International Joint Conference on Neural Networks (IJCNN), Rio de Janeiro, 8-13 July 2018, 1-8. [Google Scholar] [CrossRef]