计算机科学

计算机科学 ›› 2025, Vol. 52 ›› Issue (10): 258-265.doi: 10.11896/jsjkx.250100114

• 人工智能 • 上一篇    下一篇

基于大型语言模型文本简化的细粒度情感分析

王叶, 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:2025-01-17 修回日期:2025-05-20 出版日期:2025-10-15 发布日期:2025-10-14
  • 通讯作者: 王中卿(wangzq@suda.edu.cn)
  • 作者简介:(xiaoluhuishuohua@163.com)
  • 基金资助:
    国家自然科学基金(62076175,61976146)

Text Simplification for Aspect-based Sentiment Analysis Based on Large Language Model

WANG Ye, WANG Zhongqing   

  1. School of Computer Science and Technology,Soochow University,Suzhou,Jiangsu 215006,China
  • Received:2025-01-17 Revised:2025-05-20 Online:2025-10-15 Published:2025-10-14
  • About author:WANG Ye,born in 2002,postgraduate.Her main research interests include na-tural language processing and sentiment analysis.
    WANG Zhongqing,born in 1987,Ph.D,associate professor.His main research interests include natural language processing and sentiment analysis.
  • Supported by:
    National Natural Science Foundation of China(62076175,61976146).

PDF (PC)

2

摘要： 细粒度情感分析旨在识别句子中每个方面的情感极性。然而,现有研究大多忽视了评论文本中普遍存在的冗余信息,这些无关信息不仅增加了模型处理的复杂性,还可能导致模型无法准确捕捉原始文本中的情感元素。为解决这一问题,提出了一种将原始文本转化为简化子句的模型,以更简明的方式表达相同的情感观点。其基本思想是利用大型语言模型预识别文本中的方面词和意见词,再基于识别结果生成简化子句,并通过自我验证机制确保生成的子句满足情感一致性、相关性和简洁性。此外,所提模型结合原始文本和简化子句共同生成情感元素。在公开数据集 Restaurant和 Laptop以及 Phone上,所提模型的表现均优于现有基线模型,证明简化子句在细粒度情感分析中具有重要的作用。

关键词: 细粒度情感分析, 文本简化, 大型语言模型, 自我验证, 自然语言处理

Abstract: Aspect-based sentiment analysis aims to identify the sentiment polarity of each aspect in a sentence.However,most existing approaches overlook the redundant and irrelevant information often present in review texts,which not only complicates model processing,but also hinders accurate sentiment element extraction.To address this issue,this paper proposes a model that transforms the original text into simplified clauses,expressing the same sentiment in a more concise manner.The key idea is to leverage a large language model to pre-identify aspect and opinion terms in the text,and then generate simplified clauses based on these identified sentiment elements.A self-verification mechanism is employed to ensure the generated clause satisfy three criteria:sentiment consistency,relevance,and conciseness.Furthermore,the model jointly uses both the original text and the simplified clauses to generate sentiment elements.Experimental results on public datasets—Restaurant,Laptop,and Phone,demonstrate that the model outperforms existing baselines,highlighting the significance of simplified clauses in aspect-based sentiment analysis.

Key words: Aspect-based sentiment analysis,Text simplification, Large language model,Self-validation,Natural language proces-sing

中图分类号: 

  • TP391
[1]LIU P,JOTY S,MENG H.Fine-grained opinion mining with recurrent neural networks and word embeddings[C]//Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing.2015:1433-1443.
[2]XU H,LIU B,SHU L,et al.Double embeddings and CNN-based sequence labeling for aspect extraction[J].arXiv:1805.04601,2018.
[3]ZHOU X,WAN X,XIAO J.Representation learning for aspect category detection in online reviews[C]//Proceedings of the AAAI Conference on Artificial Intelligence.2015.
[4]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.
[5]WAN H,YANG Y,DU J,et al.Target-aspect-sentiment jointdetection for aspect-based sentiment analysis[C]//Proceedings of the AAAI Conference on Artificial Intelligence.2020:9122-9129.
[6]ZHANG W,LI X,DENG Y,et al.A survey on aspect-basedsentiment analysis:Tasks,methods,and challenges[J].IEEE Transactions on Knowledge and Data Engineering,2022,35(11):11019-11038.
[7]YAN H,DAI J,QIU X,et al.A unified generative framework for aspect-based sentiment analysis[J].arXiv:2106.04300,2021.
[8]ZHANG W,LI X,DENG Y,et al.Towards generative aspect-based sentiment analysis[C]//ACL 2021.2021.
[9]LIU J,TENG Z,CUIL,et al.Solving aspect category sentiment analysis as a text generation task[J].arXiv:2110.07310,2021.
[10]ZHANG W,DENG Y,LI X,et al.Aspect sentiment quad prediction as paraphrase generation[J].arXiv:2110.00796,2021.
[11]LU Y,LIU Q,DAI D,et al.Unified structure generation foruniversal information extraction[J].arXiv:2203.12277,2022
[12]BAO X,WANG Z,JIANG X,et al.Aspect-based SentimentAnalysis with Opinion Tree Generation[C]//IJCAI.2022:4044-4050.
[13]TANG D,QIN B,FENG X,et al.Effective LSTMs for target-dependent sentiment classification[J].arXiv:1512.01100,2015.
[14]BU J,REN L,ZHENG S,et al.ASAP:A Chinese review dataset towards aspect category sentiment analysis and rating prediction[J].arXiv:2103.06605,2021.
[15]CHEN Z,QIAN T.Relation-aware collaborative learning for unified aspect-based sentiment analysis[C]//Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics.2020:3685-3694.
[16]ZHANG M,QIAN T.Convolution over hierarchical syntacticand lexical graphs for aspect level sentiment analysis[C]//Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing(EMNLP).2020:3540-3549.
[17]GOU Z,GUO Q,YANG Y.Mvp:Multi-view prompting im-proves aspect sentiment tuple prediction[J].arXiv:2305.12627,2023.
[18]YU J,GUO Y,RUAN Q M.Aspect-level sentiment triple extraction based on set prediction[J].Journal of Chinese Information Processing,2024,38(8):147-157.
[19]WANG Q,DING K,LUO X,et al.Improving in-context learning via sequentially selection and preference alignment for few-shot aspect-based sentiment analysis[C]//Proceedings of the 47th International ACM SIGIR Conference on Research and Development in Information Retrieval.2024:2462-2466.
[20]VACAREANU R,VARIA S,HALDER K,et al.A Weak Supervision Approach for Few-Shot Aspect Based Sentiment[J].arXiv:2305.11979,2023.
[21]GOODING S.On the ethical considerations of text simplification[J].arXiv:2204.09565,2022.
[22]BIRAN O,BRODY S,ELHADAD N.Putting it simply:a context-aware approach to lexical simplification[C]//Proceedings of the 49th Annual Meeting of the Association for Computa-tional Linguistics:Human Language Technologies.2011:496-501.
[23]GOODING S,KOCHMAR E.Recursive context-aware lexicalsimplification[C]//Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing(EMNLP-IJCNLP).2019:4853-4863.
[24]SIDDHARTHAN A.Syntactic simplification and text cohesion[J].Research on Language and Computation,2006,4(1):77-109.
[25]ALUÍSIO S,GASPERIN C.Fostering digital inclusion and accessibility:the PorSimples project for simplification of Portuguese texts[C]//Proceedings of the NAACL HLT 2010 Young Investigators Workshop on Computational Approaches to Languages of the Americas.2010:46-53.
[26]BURSTEIN J,HORBACH A,KOCHMAR E,et al.Proceedings of the 16th Workshop on Innovative Use of NLP for Building Educational Applications[C]//Proceedings of the 16th Workshop on Innovative Use of NLP for Building Educational Applications.2021.
[27]DEVARAJ A,SHEFFIELD W,WALLACE B C,et al.Evaluating factuality in text simplification[C]//Proceedings of the conference.Association for Computational Linguistics.Meeting.2022,2022:7331.
[28]XIAO Z H,CHENG M M,GONG J F,et al.Research on Chinese vocabulary simplification based on hint fine-tuning[J].Journal of Chinese Information Processing,2024,38(8):34-43.
[29]SUN R,XU W,WAN X.Teaching the pre-trained model to generate simple texts for text simplification[J].arXiv:2305.12463,2023.
[30]ANSCHÜTZ M,OEHMS J,WIMMER T,et al.Language mo-dels for German text simplification:Overcoming parallel data scarcity through style-specific pre-training[J].arXiv:2305.12908,2023.
[31]TRIENES J,JOSEPH S,SCHLÖTTERER J,et al.Info-LossQA:Characterizing and recovering information loss in text simplification[J].arXiv:2401.16475,2024.
[32]JIANG X,YOU P,CHEN C,et al.Exploring scope detection for aspect-based sentiment analysis[J].IEEE/ACM Transactions on Audio,Speech,and Language Processing,2023,32:83-94.
[33]LIU D,NASSERELDINE A,YANG Z,et al.Large languagemodels have intrinsic self-correction ability[J].arXiv:2406.15673,2024.
[34]FLESCH R.A new readability yardstick[J].Journal of applied psychology,1948,32(3):221.
[35]VASWANI A,SHAZEER N,PARMAR N,et al.Attention isall you need[C]//Proceedings of the 31st International Confe-rence on Neural Information Processing Systems.2017:6000-6010.
[36]CAI H,XIA R,YU J.Aspect-category-opinion-sentiment quadruple extraction with implicit aspects and opinions[C]//Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing.2021:340-350.
[37]ZHOU J,YANG H,HE Y,et al.A unified one-step solution for aspect sentiment quad prediction[J].arXiv:2306.04152,2023.
[38]XU L,LI H,LU W,et al.Position-aware tagging for aspect sentiment triplet extraction[J].arXiv:2010.02609,2020.
[39]PEPER J J,WANG L.Generative aspect-based sentiment analysis with contrastive learning and expressive structure[J].arXiv:2211.07743,2022.
[40]ZHANG W,ZHANG X,CUI S,et al.Adaptive data augmentation for aspect sentiment quad prediction[C]//ICASSP 2024-2024 IEEE International Conference on Acoustics,Speech and Signal Processing(ICASSP).IEEE,2024:11176-11180.
[41]MA T,WANG Z,ZHOU G.Transition-based Opinion Generation for Aspect-based Sentiment Analysis[C]//Findings of the Association for Computational Linguistics ACL 2024.2024:3078-3087.
[42]BROWN T,MANN B,RYDER N,et al.Language models arefew-shot learners[J].Advances in Neural Information Proces-sing Systems,2020,33:1877-1901.
[43]TOUVRON H,MARTIN L,STONE K,et al.LLaMA 2:Open foundation and fine-tuned chat models[J].arXiv:2307.09288,2023.
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