计算机科学 ›› 2025, Vol. 52 ›› Issue (11): 223-229.doi: 10.11896/jsjkx.250500054
廖进超, 杨卫哲, 秦永彬, 黄瑞章, 陈艳平, 周裕林
LIAO Jinchao, YANG Weizhe, QIN Yongbin, HUANG Ruizhang, CHEN Yanping, ZHOU Yulin
摘要： 裁判文书自动生成是智慧法院建设中的关键任务之一,旨在提高司法效率与文书质量。由于大模型在司法审判认知上存在盲区,难以理解审理机制和文书规范,导致生成文书在逻辑一致性和结构合理性上存在不足。针对以上问题,提出了一种基于审判逻辑的裁判文书生成方法,利用大语言模型模拟审判推理过程,分阶段生成裁判文书。首先,使用法律要素填充预设模板以描述“基本案情”;其次,对事实与证据进行分析对齐得到“审理事实”;最后,结合知识库检索相关法条生成“法院判决”,并进行拼接生成完整的文书。实验结果表明,相较于基线模型,所提方法在真实案件卷宗数据上的F1值,在ROUGE-1,ROUGE-2和ROUGE-L方面分别提升了6.03,6.56和7.98个百分点,验证了所提方法的有效性。
中图分类号:
| [1]WU J.Legal document system reform under the background of intelligent justice [J].South China Sea Jurisprudence,2020,4(3):1-5. [2]Supreme People's Court.Supreme People's Court Re-leasesKey Data on Judicial Adjudication Work for 2024 [EB/OL].(2024-01-26) [2025-03-01].https://www.chinacourt.org/article/detail/2025/01/id/8686406.shtml. [3]SUTSKEVER I,VINYALS O,LE Q V.Sequence to Sequence Learning with Neural Networks [C]//Proceedings of the 28th International Conference on Neural Information Processing Systems(NIPS'14).MIT Press,2014:3104-3112. [4]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. [5]DUAN M Q.Basic Trial Procedure for First-Instance Ordinary Criminal Cases [EB/OL].(2015-11-02) [2025-03-01].https://www.66law.cn/domainblog/74744.aspx. [6]LEWIS P,PEREZ E,PIKTUS A,et al.Retrieval-augmentedgeneration for knowledge-intensive nlp tasks[J].Advances in Neural Information Processing Systems,2020,33:9459-9474. [7]DE NOVAIS E M,DIAS T T,PARABONI I.Improved TextGeneration Using N-gram Statistics [C]//12th Ibero-American Conference on Artificial Intelligence(IBERAMIA).Springer-Verlag,2010:316-325. [8]CHRISTIAN H,AGUS M P,SUHARTONO D.Single document automatic text summarization using term frequen-cy-inverse document frequency(TF-IDF)[J].ComTech:Computer,Mathematics & Engineering Applications,2016,7(4):285-294. [9]PROUDIAN D,POLLARD C.Parsing Head-Driven PhraseStructure Grammar [C]//Proceedings of the 23rd Annual Mee-ting of the Association for Computational Linguistics.1985:167-171. [10]HU H J,LIAO M F,MAO W M,et al.Variational Auto-en-coder for Text Generation [C]//Proceedings of the 5th IEEE International Conference on Information Technology and Mechatronics Engineering(ITOEC).2020:595-598. [11]LIN Z H,GONG Y Y,SHEN Y L,et al.Text Generation with Diffusion Language Models:A Pre-training Approach with Continuous Paragraph Denoise [C]//International Conference on Machine Learning.2023:21051-21064. [12]WANG Z,HE W,WU H,et al.Chinese Poetry Generation withPlanning Based Neural Network [C]//Proceedings of the 26th International Conference on Computational Linguistics(COLING).2016:1051-1060. [13]RAHMAN M M,SIDDIQUI F H.Multi-layered attentionalpeephole convolutional LSTM for abstractive text summarization[J].Etri Journal,2021,43(2):288-298. [14]YANG P C,LI L,LUO F L,et al.Enhancing Topic-to-Essay Generation with External Commonsense Knowledge [C]//Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics.2019:2002-2012. [15]PEI B S,L X,HU K Q,et al.Generation of judicial text abstracts based on knowledge enhancement pre-training model [J].Science Technology and Engineering,2024,24(20):8587-8597. [16]ALOKLA A,GAD W,NAZIH W,et al.Pseudocode generation from source code using the bart model[J].Mathematics,2022,10(21):3967. [17]ABADI V N M,GHASEMIAN F.Enhancing Persian text summarization through a three-phase fine-tuning and reinforcement learning approach with the mT5 transformer model[J].Scienti-fic Reports,2025,15(1):80. [18]ALI S R,DOBBS T D,HUTCHINGS H A,et al.Using ChatGPT to write patient clinic letters[J].The Lancet Digital Health,2023,5(4):e179-e181. [19]WANG Y,ZHOU Q,LEDO D.StoryVerse:Towards Co-authoring Dynamic Plot with LLM-based Character Simulation via Narrative Planning [C]//Proceedings of the 19th International Conference on the Foundations of Digital Games.2024:1-4. [20]LI H Z,WANG H Y,SUN X,et al.Prompt-guided Generation of Structured Chest X-ray Report Using a Pre-trained LLM [C]//Proceedings of the IEEE International Conference on Multimedia and Expo(ICME).2024:1-6. [21]XIAO S T,LIU Z,ZHANG P T,et al.C-Pack:Packed Resources for General Chinese Embeddings [C]//Proceedings of the 47th International ACM SIGIR Conference on Research and Development in Information Retrieval.2024:641-649. [22]LIN C.Rouge:A Package for Automatic Evaluation of Summaries [C]//Proceedings of the Workshop on Text Summarization Branches Out.2004:74-81. [23]MIN S,LYU X,HOLTZMAN A,et al.Rethinking the Role ofDemonstrations:What Makes In-Context Learning Work?[C]//Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing.2022:11048-11064. [24]MIHALCEA R,TARAU P.TextRank:Bringing Order intoText [C]//Proceedings of the 2004 Conference on Empirical Methods in Natural Language Processing.2004:404-411. [25]SHI Y S,MENG J,WANG J.Seq2Seq Model with RNN Attention for Abstractive Summarization [C]//Proceedings of the 2019 International Conference on Artificial Intelligence and Computer Science.2019:348-353. [26]LEWIS M,LIU Y,GOYAL N,et al.BART:Denoising Se-quence-to-Sequence Pre-training for Natural Language Generation,Translation,and Comprehension [C]//58th Annual Mee-ting of the Association for Computational Linguistics.2020:7871-7880. [27]SHAO Y F,GENG Z C,LIU Y T,et al.Cpt:A pre-trained unbalanced transformer for both chinese language understanding and generation[J].Science China Information Sciences,2024,67(5):152102. [28]DU Z X,QIAN Y J,LIU X,et al.GLM:General Language Mo-del Pre-training with Autoregressive Blank Infilling [C]//60th Annual Meeting of the Association for Computational Linguistics.2022:320-335. [29]GRATTAFIORI A,DUBEY A,JAUHRI A,et al.The llama 3 herd of models[J].arXiv:2407.21783,2024. [30]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(1):5485-5551. |
|
||
首页