Page 287 - 《软件学报》2025年第5期

琚江舟 等: 多粒度单元格对比的文本和表格数值问答模型 2187

[30] Vaswani A, Shazeer N, Parmar N, Uszkoreit J, Jones L, Gomez AN, Kaiser Ł, Polosukhin I. Attention is all you need. In: Proc. of the
31st Int’l Conf. on Neural Information Processing Systems. Long Beach: Curran Associates Inc., 2017. 6000–6010.
[31] Chen XY, Liang C, Yu AW, Zhou D, Song D, Le QV. NEURAL symbolic reader: Scalable integration of distributed and symbolic
representations for reading comprehension. In: Proc. of the 8th Int’l Conf. on Learning Representations. Addis Ababa, 2020.
[32] Zhu FB, Lei WQ, Huang YC, Wang C, Zhang S, Lv JC, Feng FL, Chua TS. TAT-QA: A question answering benchmark on a hybrid of
tabular and textual content in finance. In: Proc. of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th
Int’l Joint Conf. on Natural Language Processing (Vol. 1: Long Papers). Association for Computational Linguistics, 2021. 3277–3287.
[doi: 10.18653/v1/2021.acl-long.254]
[33] Ran Q, Lin YK, Li P, Zhou J, Liu ZY. NumNet: Machine reading comprehension with numerical reasoning. In: Proc. of the 2019 Conf.
on Empirical Methods in Natural Language Processing and the 9th Int’l Joint Conf. on Natural Language Processing (EMNLP-IJCNLP).
Hong Kong: Association for Computational Linguistics, 2019. 2474–2484. [doi: 10.18653/v1/D19-1251]
[34] Araci D. FinBERT: Financial sentiment analysis with pre-trained language models. arXiv:1908.10063, 2019.
[35] Wei J, Wang XZ, Schuurmans D, Bosma M, Ichter B, Xia F, Chi E, Le Q, Zhou D. Chain-of-thought prompting elicits reasoning in large
language models. arXiv:2201.11903, 2023.
[36] Touvron H, Martin L, Stone K, et al. Llama 2: Open foundation and fine-tuned chat models. arXiv:2307.09288, 2023.
[37] Introducing ChatGPT. 2022. https://openai.com/blog/chatgpt
[38] Liu YH, Ott M, Goyal N, Du JF, Joshi M, Chen DQ, Levy O, Lewis M, Zettlemoyer L, Stoyanov V. RoBERTa: A robustly optimized
BERT pretraining approach. arXiv:1907.11692, 2019.
[39] Loshchilov I, Hutter F. Decoupled weight decay regularization. arXiv:1711.05101, 2019.
[40] Srivastava P, Malik M, Gupta V, Ganu T, Roth D. Evaluating LLMs’ mathematical reasoning in financial document question answering.
arXiv:2402.11194, 2024.
[41] Wang F, Liu HP. Understanding the behaviour of contrastive loss. In: Proc. of the 2021 IEEE/CVF Conf. on Computer Vision and Pattern
Recognition (CVPR). Nashville: IEEE, 2021. 2495–2504. [doi: 10.1109/CVPR46437.2021.00252]

附中文参考文献:
[1] 游斓, 周雅倩, 黄萱菁, 吴立德. 基于最大熵模型的 QA 系统置信度评分算法. 软件学报, 2005, 16(8): 1407–1414. http://www.jos.org.
cn/1000-9825/16/1407.htm [doi: 10.1360/jos161407]
[2] 乔少杰, 杨国平, 于泳, 韩楠, 覃晓, 屈露露, 冉黎琼, 李贺. QA-KGNet: 一种语言模型驱动的知识图谱问答模型. 软件学报, 2023,
34(10): 4584–4600. http://www.jos.org.cn/1000-9825/6882.htm [doi: 10.13328/j.cnki.jos.006882]
[25] 陈跃鹤, 贾永辉, 谈川源, 陈文亮, 张民. 基于知识图谱全局和局部特征的复杂问答方法. 软件学报, 2023, 34(12): 5614–5628. http://
www.jos.org.cn/1000-9825/6799.htm [doi: 10.13328/j.cnki.jos.006799]

琚江舟(1991－), 男, 博士生, 主要研究领域为问 陈宇飞(2000－), 男, 硕士生, 主要研究领域为问
答系统. 答系统.

毛云麟(1999－), 男, 硕士生, 主要研究领域为问 戴新宇(1979－), 男, 博士, 教授, 博士生导师,

答系统. CCF 专业会员, 主要研究领域为自然语言处理,

人机对话交互, 人工智能应用.

吴震(1993－), 男, 博士, 助理教授, 主要研究领 陈家骏(1963－), 男, 博士, 教授, 博士生导师,

域为情感分析, 观点挖掘, 情感生成, 迁移学习. CCF 专业会员, 主要研究领域为自然语言处理,

机器翻译, 程序设计语言.