| 789 | 8 | 724 |
| 下载次数 | 被引频次 | 阅读次数 |
信用评级作为债券市场的重要基础设施,其行业的健康发展对债券市场高质量发展具有重要意义。人工智能技术特别是预训练语言模型的提出和大型语言模型技术的快速发展,对信用评级行业健康发展带来重大机遇。新兴人工智能技术在非结构化数据处理、信用模型构建、信用分析等方面提供了全新的技术来源,促使传统信用评级流程和评级方式发生变革;通过数智赋能提升信用评级预警能力和评级质效,催生评级大数据、评级数字原生等信用评级新生态。但应用人工智能技术也面临技术成熟度、技术透明度和可信赖度等方面的挑战,需要在工程层面通过技术创新来解决诸多问题。通过对人工智能创新评级数据处理、评级分析、评级报告撰写等多个具体场景中的实践和探索分析,探讨人工智能赋能信用评级高质量发展的优势和前景,为推动信用评级迈入数智时代提供有效路径。
Abstract:Credit rating, as an important infrastructure of the bond market, plays a significant role in the high-quality development of the bond market. The emergence of artificial intelligence technology, especially Pre-trained Language Model, and the rapid development of Large Language Model technology have brought major opportunities for the credit rating industry. Emerging artificial intelligence technologies provide new technical sources for non-structured data processing, credit modeling process, and credit analysis. the change is happening in traditional rating processes and methods. The credit rating industry is continuously enhancing its warning capability,rating quality and efficiency through digital empowerment, while also giving rise to new credit rating ecosystems with big data and digital natives. Credit rating agencies also face challenges in terms of the maturity, transparency, and reliability of artificial intelligence technology when applying it in practice. These challenges need to be addressed through technological innovation at the engineering level. By analyzing the practice and exploration of artificial intelligence in multiple specific scenarios such as rating data processing, rating analysis, and rating report writing, this paper discusses the advantages and prospects of artificial intelligence in empowering the high-quality development of credit ratings,providing an effective path for promoting the credit rating industry into the era of digit and intelligence.
[1]傅小康,邹江波,琚春华.信用评估研究综述与演化分析[J].征信,2021(6):7-15+21.
[2]Dell M. Deep Learning for Economists[R/OL].(2024-08)[2024-10-15]. https://www.nber.org/pagers/w32768.
[3]罗锦钊,孙玉龙,钱增志,等.人工智能大模型综述及展望[J].无线电工程,2023(11):2461-2472.
[4]Hadji I.,Wildes R. P. What do we understand about convolutional networks?[J]. arxiv preprint arxiv:1803.08834,2018.
[5]Lewis P.,Perez E.,Piktus A.,et al. Retrieval-augmented generation for knowledge-intensive NLP Tasks[J]. Advances in Neural Information Processing Systems,2020(33):9459-9474.
[6]Wei J.,Wang X.,Schuurmans D.,et al. Chain-of-thought Prompting Elicits Reasoning in Large Language Models[J].Advances in neural information processing systems,2022(35):24824-24837.
[7]周涛,李鑫,周俊临,等.大模型智能体:概念、前沿和产业实践[J].电子科技大学学报(社科版),2024(4):57-62.
[8]董昀,何浩钦.中国的数字金融发展:理论机制、现实背景与政策建议[J].东方论坛-青岛大学学报(社会科学版),2024(5):59-68.
[9]Cheng X.,Liu S.,Sun X.,et al. Combating Emerging Financial Risks in the Big Data Era:A Perspective Review[J]. Fundamental Research, 2021,1(5):595-606.
[10]李煜华,舒慧珊,向子威.数字原生企业与非原生企业数字化转型组态路径研究:基于“技术-组织-环境”理论框架[J].软科学,2023(7):58-65.
[11]李国权.“人工智能+信用治理”:逻辑理路、边界约束及实践路向[J].征信,2024(8):20-26.
[12]Zhou A.,Wang K.,Lu Z.,et al. Solving Challenging Math Word Problems Using GPT-4 Code Interpreter with Code-based Self-Verification[J].arxiv preprint arxiv:2308.07921,2023.
(1)数据来源:根据Wind数据库数据统计所得。
(1)参见《十四届全国人大常委会专题讲座第十讲讲稿——人工智能与智能计算的发展》,https://www.seiot.com.cn/detail/23030.htm。
(1)数据来源:根据Wind数据库数据统计所得。
基本信息:
中图分类号:TP18;F832.51
引用信息:
[1]赵维波.人工智能在信用评级中的实践和探索[J].征信,2025,43(02):39-47.
2025-02-26
2025-02-26
2025-02-26