大语言模型在决策咨询工作中的应用解析Application Analysis of Large Language Models in Decision Consulting Work
黄吉海,周雷,王延飞
摘要(Abstract)：
[目的/意义]大语言模型在决策咨询领域具有较大应用潜力，分析其面临的精准性与可靠性问题，提出针对性解决方案具有重要意义。[方法/过程]本文立足当前大语言模型的特点，分析探讨了其在决策咨询应用中暴露出的模型幻觉、持续学习能力不足、多模态输入输出融合困难、可解释性和价值观对齐等问题。[结果/结论]研究提出了构建“知识增强+专家验证”、优化增量训练、强化符号逻辑与大模型融合、打造全息信息管道以及建设语料约束体系的五维应对框架，推动大语言模型赋能决策咨询。
关键词(KeyWords)： 大语言模型;决策咨询;应用解析;模型幻觉;持续学习
基金项目(Foundation):
作者(Author): 黄吉海,周雷,王延飞
DOI: 10.19318/j.cnki.issn.2096-1634.2025.04.05
参考文献(References)：
- [1]OUYANG L,WU J,XU J,et al.Training language models to follow instructions with human feedback[EB/OL].(2022-03-04)[2025-03-10].https://arxiv.org/pdf/2203.02155.
- [2]DeepSeek-AI.DeepSeek-R1:Incentivizing reasoning capability in LLMs via reinforcement learning[EB/OL].(2025-01-22)[2025-03-10].https://arxiv.org/pdf/2501.12948.
- [3]观察者网.DeepSeek在全球140个市场下载量排名第一，印度新增用户占比最多[EB/OL].(2025-02-01)[2025-03-24].https://www.guancha.cn/industry-science/2025_02_01_763824.shtml.
- [4]魏宏程，杨建林.大语言模型+检索增强方法的关键技术及其在情报任务中的应用流程[J].情报理论与实践，2025,48(3):178-188,206.
- [5]李阳，孙建军.大模型对情报学发展的影响思考[J].情报学报，2025,44(2):246-256.
- [6]WANG C J,SENNRICH R.On exposure bias,hallucination and domain shift in neural machine translation[C]// Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics.Stroudsburg:ACL,2020:3544-3552.
- [7]YU T,ZHANG Y F,FU C Y,et al.Aligning multimodal LLM with human preference:A survey[EB/OL].(2025-03-23)[2025-03-27].https://arxiv.org/pdf/2503.14504.
- [8]DeepSeek-AI.DeepSeek-V3 technical report[EB/OL].(2025-02-18)[2025-03-10].https://arxiv.org/pdf/2412.19437.
- [9]LOSING V,HAMMER B,WERSING H.Incremental on-line learning:A review and comparison of state of the art algorithms[J].Neurocomputing,2018(275):1261-1274.
- [10]成素梅.人工智能研究的范式转换及其发展前景[J].哲学动态，2017(12):15-21.
- [11]王广赞，易显飞.人工智能研究的三大流派：比较与启示[J].长沙理工大学学报(社会科学版),2018,33(4):1-6.
- [12]XU J D,FEI H,PAN L M,et al.Faithful logical reasoning via symbolic chain-of-thought[EB/OL].(2024-06-11)[2025-03-10].https://arxiv.org/pdf/2405.18357.
- [13]SINHA S,PRABHU A,KUMARAGURU P,et al.Wu’s method can boost symbolic AI to rival silver medalists and AlphaGeometry to outperform gold medalists at IMO geometry[J].(2024-04-11)[2025-03-10].https://arxiv.org/pdf/2404.06405.