融合简化句法信息的端到端方面级情感分析

doi:10.3778/j.issn.1002-8331.2206-0376

摘要/Abstract

摘要： 端到端方面级情感分析（E2E-ABSA）通过联合标签的方式识别评论中的方面术语，并对其进行情感极性分类。现有模型关注于改进特征编码器和解码器来提升性能。对于该任务而言，方面术语通常以组块的形式出现，且方面术语和观点术语具有句法关联，有效利用上述特性将有利于E2E-ABSA。提出了一种融合简化句法信息的E2E-ABSA模型。通过常用的句法分析获得评论的原始句法依赖树。同时，制定一组规则对句法依赖树进行重构，以获得简洁的句法信息。利用注意力机制将重构的句法信息融合到模型中，获得方面术语和观点术语的相关特征，从而更好地完成方面级情感分析。在两个公共评论数据集上进行验证，实验结果证明了该模型的有效性，且消融实验证明简化的句法信息确实有利于端到端方面级情感分析任务的完成。

关键词: 方面级情感分析, 句法, 端到端, 方面术语, 注意力机制

Abstract: End-to-end aspect-based sentiment analysis（E2E-ABSA） recognizes the aspect terms from reviews and classifies their sentiment polarities by an unified tagging scheme. Existing models mainly focus on improving feature encoder and decoder to improve performance. Existing models focus on improving feature encoder and decoder to improve performance. For this task, aspect terms usually appear in the form of chunks, and aspect terms and opinion terms are syntactically related, effective use of above characteristic would benefit E2E-ABSA. An E2E-ABSA model is proposed to incorporate simplified syntactic information. Firstly, it gets the original dependency tree of a review by a common syntactic parsing. At the same time, it makes a set of rules to reconstruct the original dependency tree to obtain concise syntactic information. Then it uses attention mechanism to integrate the reconstructed syntactic information into the model, and obtains the correlation aspect term and opinion term features, so as to better complete aspect level sentimental analysis. Experimental results on two public review datasets show the effectiveness of the model, and ablation experiments show that the simplified syntactic information is indeed conducive to the completion E2E-ABSA.

Key words: based-aspect sentiment analysis, syntactic, end-to-end, aspect terms, attention mechanism

张济群, 张名芳, 郭军军, 相艳. 融合简化句法信息的端到端方面级情感分析[J]. 计算机工程与应用, 2023, 59(20): 129-137.

ZHANG Jiqun, ZHANG Mingfang, GUO Junjun, XIANG Yan. End-to-End Based-Aspect Sentiment Analysis Incorporating Simplified Syntactic Information[J]. Computer Engineering and Applications, 2023, 59(20): 129-137.

参考文献

[1] ARWA S，MOSTAFA E，OSAMA A A.Arabic aspect extraction based on stacked contextualized embedding with deep learning[J].IEEE Access，2022，10：30526-30535.
[2] AYAL K，OREN P，DANIEL K，et al.Opinion-based relational pivoting for cross-domain aspect term extraction[C]//Proceedings of the 12th Workshop on Computational Approaches to Subjectivity，Sentiment & Social Media Analysis，Dublin，2022：104-112.
[3] 赵丽华，王春立，初钰凤.基于注意力双层BiReGU模型的方面术语提取方法[J].计算机工程与应用，2021，57（22）：160-165.
ZHAO L H，WANG C L，CHU Y F.Attention-based double BiReGU model for aspect term extraction[J].Computer Engineering and Applications，2021，57（22）：160-165.
[4] WANG Q L，WEN Z Y，ZHAO Q，et al.Progressive self-training with discriminator for aspect term extraction[C]//Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing，2021：257-268.
[5] XUE W，LI T.Aspect based sentiment analysis with gated convolutional networks[C]//Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics，Melbourne，2018：2514-2523.
[6] LI X，BING L D，LI P J，et al.A unified model for opinion target extraction and target sentiment prediction[C]//The Thirty-First Innovative Applications of Artificial Intelligence Conference，Hawaii，2019：6714-6721.
[7] LI X，BING L D，ZHANG W X，et al.Exploiting BERT for end-to-end aspect-based sentiment analysis[C]//Proceedings of the 5th Workshop on Noisy User-generated Text，Hong Kong，2019：34-41.
[8] 杨春霞，宋金剑，姚思诚.面向方面级情感分析的加权依存树卷积网络[J].中文信息学报，2022，36（5）：125-132.
YANG C X，SONG J J，YAO S C.A weighted dependency tree convolutional networks for aspect-based sentiment analysis[J].Journal of Chinese Information Processing，2022，36（5）：125-132.
[9] DAN J，JOYCE C，NATALIE S，et al.Embarrassingly simple unsupervised aspect extraction[C]//Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics，2020：3182-3187.
[10] WEI Z K，HONG Y，ZOU B W，et al.Don’t eclipse your arts due to small discrepancies：boundary repositioning with a pointer network for aspect extraction[C]//Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics，2020：3678-3684.
[11] PENG H Y，XU L，BING L D，et al.Knowing what，how and why：a near complete solution for aspect-based sentiment analysis[C]//The Thirty-Second Innovative Applications of Artificial Intelligence Conference，New York，2020：8600-8607.
[12] MAO Y，YI S，YU C，et al.A joint training dual-MRC framework for aspect based sentiment analysis[C]//Thirty-Third Conference on Innovative Applications of Artificial Intelligence，2021：13543-13551.
[13] SCHMITT M，STEINHEBER S，SCHREIBER K，et al.Joint aspect and polarity classification for aspect-based sentiment analysis with end-to-end neural networks[C]//Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing，Brussels，2018：1109-1114.
[14] LI X，BING L D，LI P J，et al.Aspect term extraction with history attention and selective transformation[C]//Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence，Stockholm，2018：4194-4200.
[15] MA D H，LI S J，WU F Z.Exploring sequence-to-sequence learning in aspect term extraction[C]//Proceedings of the 57th Conference of the Association for Computational Linguistics，Florence，2019：3538-3547.
[16] LI K，CHEN C B，QUAN X J，et al.Conditional augmentation for aspect term extraction via masked sequence-to-sequence generation[C]//Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics，2020：7056-7066.
[17] MA D H，LI S J，ZHANG X D.Interactive attention networks for aspect-level sentiment classification[C]//Proceedings of the Twenty-Sixth International Joint Conference on Artificial Intelligence，Melbourne，2017：4068-4074.
[18] YI T，LUU A T，SIU C，et al.Learning to attend via word-aspect associative fusion for aspect-based sentiment analysis[C]//Proceedings of the Thirty-Second AAAI Conference on Artificial Intelligence，Louisiana，2018：5956-5963.
[19] LI X，BING L，LAM W.Transformation networks for target-oriented sentiment classification[C]//Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics，Melbourne，2018：946-956.
[20] KE Z X，LIU B，XU H，et al.CLASSIC：continual and contrastive learning of aspect sentiment classification tasks[C]//Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing，2021：6871-6883.
[21] CHEN P，SUN Z Q，BING L D.Recurrent attention network on memory for aspect sentiment analysis[C]//Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing，Copenhagen，2017：452-461.
[22] MA D H，LI S J，WANG H F.Joint learning for targeted sentiment analysis[C]//Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing，Brussels，Belgium，2018：4737-4742.
[23] KLINGER R，CIMIANO P.Bi-directional inter-dependencies of subjective expressions and targets and their value for a joint model[C]//Proceedings of the 51st Annual Meeting of the Association for Computational Linguistics，Sofia，2013：848-854.
[24] YANG B S，CLAIRE C.Joint inference for fine-grained opinion extraction[C]//Proceedings of the 51st Annual Meeting of the Association for Computational Linguistics，Sofia，2013：1640-1649.
[25] LUO H S，LI T R，LIU B，et al.DOER：dual cross-shared RNN for aspect term-polarity co-extraction[C]//Proceedings of the 57th Conference of the Association for Computational Linguistics，Florence，2019：591-601.
[26] HE R D，WEE S L，HWEE T N，et al.An interactive multi-task learning network for end-to-end aspect-based sentiment analysis[C]//Proceedings of the 57th Conference of the Association for Computational Linguistics，Florence，2019：504-515.
[27] LIU S，LI W，WU Y F，et al.Jointly modeling aspect and sentiment with dynamic heterogeneous graph neural networks[J].arXiv：2004.06427，2020.
[28] ASHISH V，NOAM S，NIKI P，et al.Attention is all you need[C]//Advances in Neural Information Processing Systems 30：Annual Conference on Neural Information Processing Systems 2017，Long Beach，2017：5998-6008.
[29] MARIA P，DIMITRIS G，HARIS P，et al.SemEval-2016 task 5：aspect based sentiment analysis[C]//Proceedings of the 10th International Workshop on Semantic Evaluation，San Diego，2016：19-30.
[30] MARIA P，DIMITRIS G，HARIS P，et al.SemEval-2015 task 12：aspect based sentiment analysis[C]//Proceedings of the 9th International Workshop on Semantic Evaluation，Colorado，2015：486-495.
[31] MARIA P，DIMITRIS G，JOHN P，et al.SemEval-2014 task 4：aspect based sentiment analysis[C]//Proceedings of the 8th International Workshop on Semantic Evaluation，Dublin，2014：27-35.
[32] JACOB D，KENTON L.BERT：pre-training of deep bidirectional transformers for language understanding[C]//Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics：Human Language Technologies，Minneapolis，2019：4171-4186.
[33] MARGARET M，JACQUI A，THERESA W，et al.Open domain targeted sentiment[C]//Proceedings of the 2013 Conference on Empirical Methods in Natural Language Processing，Seattle，2013：1643-1654.
[34] MA X Z，EDUARD H H.End-to-end sequence labeling via bi-directional LSTM-CNNs-CRF[C]//Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics，Berlin，2016：1064-1074.
[35] BIE Y，YANG Y.A multitask multi-view neural network for end-to-end aspect-based sentiment analysis[J].Big Data Mining and Analytics，2021，4（3）：195-207.
[36] LU T，XIANG Y，ZHANG L，et al.Sentence constituent-aware attention mechanism for end-to-end aspect-based sentiment analysis[J].Multimedia Tools and Applications，2022，81：15333-15348.