计算机科学

计算机科学 ›› 2025, Vol. 52 ›› Issue (8): 300-307.doi: 10.11896/jsjkx.240900114

• 人工智能 • 上一篇    下一篇

结合预训练模型和数据增强的跨领域属性级情感分析研究

陈舸, 王中卿   

  1. 苏州大学计算机科学与技术学院 江苏 苏州 215006
  • 收稿日期:2024-09-18 修回日期:2024-12-03 出版日期:2025-08-15 发布日期:2025-08-08
  • 通讯作者: 王中卿(wangzq＠suda.edu.cn)
  • 作者简介:(20245227045@stu.suda.edu.cn)

Cross-domain Aspect-based Sentiment Analysis Based on Pre-training Model with Data Augmentation

CHEN Ge, WANG Zhongqing   

  1. School of Computer Science and Technology,Soochow University,Soochow,Suzhou 215006,China
  • Received:2024-09-18 Revised:2024-12-03 Online:2025-08-15 Published:2025-08-08
  • About author:CHEN Ge, born in 2001, postgraduate, is a member of CCF（No.V5365G). Her main research interests include sentiment analysis and so on.
    WANG Zhongqing, born in 1987, Ph.D, associate professor. His main research interests include natural language processing, sentiment analysis and information extraction.

PDF (PC)

24

摘要： 属性级情感分析(ABSA)是一项细粒度情感分析任务,旨在识别文本中的具体属性并探测其情感倾向。针对ABSA模型因无法适应不同领域的语言风格而导致性能不佳以及目标领域缺乏标注数据的问题,提出了一种结合预训练模型的跨领域属性级情感分析方法。该方法利用预训练模型对目标领域文本进行标签生成,再利用大语言模型重新生成更具目标领域风格的自然语句,最后将生成的样本和源领域样本组合训练,以对目标领域进行预测。在SemEval语料库的restaurant和laptop数据集以及一个公开的网络服务评论数据集上进行实验,结果表明,与现有跨领域情感分析方法相比,所提方法在F1值上至少提升了5.33%,充分证明了该方法的有效性。

关键词: 跨领域情感分析, 预训练模型, T5, GPT

Abstract: Aspect-based Sentiment Analysis(ABSA) is a fine-grained sentiment analysis task,which aimes at identifying specific aspects in text and exploring their sentiment orientation.To solve the problem of poor performance of ABSA model due to its in-ability to adapt to different domain language styles and lack of labeled data in target domain,this paper proposes a cross-domain aspect-based sentiment analysis method combined with pre-trained model.The pretraining model is used to generate labels for the target domain text,and the large language model is used to regenerate natural sentences with more target domain style.Finally,the generated samples and source domain samples are combined for training to predict the target domain.This experimental results on the restaurant and laptop datasets from the SemEval corpus,as well as a publicly available Web service review dataset show that,compared to existing cross-domain sentiment analysis methods,the proposed method achieves at least a 5.33% improvement in F1 score,fully demonstrating its effectiveness.

Key words: Cross-domain ABSA, Pre-training model, T5, GPT

中图分类号: 

  • TP391
[1]HU M,LIU B.Mining and summarizing customer reviews[C]//Proceedings of the Tenth ACM SIGKDD International Confe-rence on Knowledge Discovery and Data Mining.2004:168-177.
[2]WANG W,PAN S J,DAHLMEIER D,et al.Coupled multi-layer attentions for co-extraction of aspect and opinion terms[C]//Proceedings of the AAAI Conference on Artificial Intelligence.2017:3316-3322.
[3]HU M,ZHAO S,ZHANG L,et al.CAN:Constrained attentionnetworks for multi-aspect sentiment analysis[C]//Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing.2019:4601-4610.
[4]LI X,BING L,LI P,et al.A unified model for opinion target extraction and target sentiment prediction[C]//Proceedings of the AAAI Conference on Artificial Intelligence.2019:6714-6721.
[5]PENG H,XU L,BING L,et al.Knowing what,how and why:A near complete solution for aspect-based sentiment analysis[C]//Proceedings of the AAAI Conference on Artificial Intelligence.2020:8600-8607.
[6]XU L,LI H,LU W,et al.Position-aware tagging for aspect sentiment triplet extraction[C]//Proceedings of the 2020 Confe-rence on Empirical Methods in Natural Language Processing.2020:2339-2349.
[7]WU Z,YING C,ZHAO F,et al.Grid tagging scheme for aspect-oriented fine-grained opinion extraction[C]//Findings of the Association for Computational Linguistics:EMNLP 2020.2020:2576-2585.
[8]ZHANG W,LI X,DENG Y,et al.Towards generative aspect-based sentiment analysis[C]//Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing. 2021:504-510.
[9]ZHANG W,DENG Y,LI X,et al.Aspect sentiment quad prediction as paraphrase generation[C]//Proceedings of the 2021 Conference on Empirical Methods in Natural Language Proces-sing.2021:9209-9219.
[10]BAI Y,XIE Y,LIU X,et al.BvSP:Broad-view Soft Prompting for Few-Shot Aspect Sentiment Quad Prediction[C]//Procee-dings of the 62nd Annual Meeting of the Association for Computational Linguistics.2024:8465-8482.
[11]CHEN B,OUYANG Q,LUO Y,et al.S2GSL:Incorporating Segment to Syntactic Enhanced Graph Structure Learning for Aspect-based Sentiment Analysis[C]//Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics.2024:13366-13379.
[12]JAKOB N,GUREVYCH I.Extracting opinion targets in a single and cross-domain setting with conditional random fields[C]//Proceedings of the 2010 Conference on Empirical Methods in Natural Language Processing.2010:1035-1045.
[13]CHERNYSHEVICH M,BELARUS I.Cross-domain extraction of product features using conditional random fields[C]//Proceedings of the 8th International Workshop on Semantic Evaluation(SemEval 2014).2014:309-313.
[14]DING Y,YU J,JIANG J.Recurrent neural networks with auxiliary labels for cross-domain opinion target extraction[C]//Proceedings of the AAAI Conference on Artificial Intelligence.2017:3436-3442.
[15]YU J,GONG C,XIA R.Cross-domain review generation for aspect-based sentiment analysis[C]//Findings of the Association for Computational Linguistics:ACL-IJCNLP 2021.2021:4767-4777.
[16]SU Y,ZHOU X B.Cross-domain Sentiment Analysis Based on Gradient Data Selection[J].Software guid,2023,22(5):50-56.
[17]DENG Y,ZHANG W,PAN S J,et al.Bidirectional generativeframework for cross-domain aspect-based sentiment analysis[C]//Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics.2023:12272-12285.
[18]RAFFEL C,SHAZEER N,ROBERTS A,et al.Exploring the limits of transfer learning with a unified text-to-text transformer[J].Journal of Machine Learning Research,2020,21(140):1-67.
[19]LI Z,LI X,WEI Y,et al.Transferable end-to-end aspect-based sentiment analysis with selective adversarial learning[C]//Proceedings of the 2019 Confe-rence on Empirical Methods in Natural Language Processing and the 9th International Joint Confe-renceon Natural Language Processing.2019:4590-4600.
[20]ZHOU Y,ZHU F,SONG P,et al.An adaptive hybrid framework for cross-domain aspect-based sentiment analysis[C]//Proceedings of the AAAI Conference on Artificial Intelligence.2021:14630-14637.
[21]GONG C,YU J,XIA R.Unified feature andinstance based domain adaptation for aspect-based sentiment analysis[C]//Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing.2020:7035-7045.
[22]LI J,YU J,XIA R.Generative cross-domain data augmentation for aspect and opinion co-extraction[C]//Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics:Human Language Technologies.2022:4219-4229.
[23]YU J,ZHAO Q,XIA R.Cross-domain data augmentation with domain-adaptive language modeling for aspect-based sentiment analysis[C]//Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics.2023:1456-1470.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
渝ICP备16013121号-4
地址：重庆市渝北区洪湖西路18号    邮编：401121  
电话：023-63500828（编辑部） 023-67039612（会议/专题） 023-67039623（财务/发票）
E-mail：jsjkx12@163.com
本系统由北京玛格泰克科技发展有限公司设计开发