联系我们

论文查询结果	返回搜索

论文编号：	14979
作者编号：	2320224298
上传时间：	2024/12/8 10:53:07
中文题目：	大语言模型在项目决策质量管理中的应用研究——以HS公司海外SaaS软件开发项目为例
英文题目：	Research on the Application of Large Language Models in Project Decision Quality Management ——Taking HS Company''s Overseas SaaS Software Development Project as an Example
指导老师：	石鉴
中文关键字：	大语言模型；项目决策管理；决策质量提升；人工智能应用；智能项目管理
英文关键字：	Large Language Models (LLMs)；Project Decision Management；Decision Quality Improvement；Artificial Intelligence Application；Intelligent Project Management
中文摘要：	随着项目规模和复杂性的增加，传统的项目决策管理方法逐渐显现出局限性，尤其是在处理海量信息、制定备选方案、利益相关者决策偏好沟通，以及决策执行控制等方面，传统项目决策过程逐步暴露出诸多问题，进而影响项目决策质量。大语言模型作为人工智能领域的重要技术进展，凭借其强大的自然语言处理和数据分析能力，为项目管理中的决策过程提供了新的解决思路。本研究旨在探讨如何利用大语言模型技术及其周边技术在项目决策质量管理中的应用，以提升项目决策的科学性和精确性，并探寻构建出一套系统化、规范化的决策流程。 本研究首先对项目决策管理与决策过程现状进行了梳理分析，暴露当前项目决策过程中所存在的问题及其成因。然后对大语言模型及其周边技术在各项问题中的应用路径与应用方案进行了研究，包括以下几点：1 探讨在有限时间内如何从海量信息中快速提取高质量的情报，提升决策质量；2 分析传统算法在项目决策中利益相关者决策偏好沟通的障碍，并提出改进方法；3 研究传统算法的决策过程可解释性问题，探索如何使大语言模型提供明确的决策解释；4 探讨大语言模型在项目决策执行控制过程中的应用，尤其是在动态环境中实现决策控制的优化。本研究通过对这些问题的深入分析，提出了一套以大语言模型为工具的项目决策质量管理改进方案。最后，本文将改进方案应用到HS公司具体项目中，完成了实际应用，并对应用过程与效果进行了分析。 本研究的创新点在于首次系统性地将大语言模型引入到项目决策质量管理中，并在多个决策环节中探讨其应用价值。通过应用大语言模型，本文在情报获取、决策沟通、可解释性分析及执行控制等方面提出了具体的应用方案，旨在提升项目管理中的决策质量与效率。研究结果不仅为项目管理领域提供了新的技术手段，也为大语言模型技术的发展及其在其他管理领域的应用奠定了基础，具有重要的理论和实践意义。
英文摘要：	As the scale and complexity of projects increase, traditional project decision-making management methods gradually reveal limitations, especially in handling massive amounts of information, developing alternative plans, communicating stakeholder decision preferences, and controlling decision execution. These challenges in traditional project decision processes have exposed various issues that ultimately affect decision quality. Large language models (LLMs), a significant advancement in the field of artificial intelligence, offer a new approach to decision-making in project management, thanks to their powerful natural language processing and data analysis capabilities. This study aims to explore how LLMs and related technologies can be applied to project decision quality management to enhance the scientific and precise nature of decisions, while also exploring the construction of a systematic and standardized decision-making process. This study first reviews and analyzes the current state of project decision management and decision-making processes, identifying existing problems and their causes. It then explores the application paths and solutions for LLMs and related technologies to address these problems, focusing on the following key areas: (1) how to quickly extract high-quality intelligence from massive amounts of information within limited timeframes to improve decision quality; (2) analyzing the communication barriers in stakeholder decision preferences posed by traditional algorithms in project decision-making and proposing improvements; (3) investigating the issue of decision process interpretability in traditional algorithms and exploring how LLMs can provide clear decision explanations; and (4) exploring the application of LLMs in decision execution control, particularly optimizing decision control in dynamic environments. Through in-depth analysis of these issues, the study proposes an improvement plan for project decision quality management using LLMs as a tool. Finally, the study applies this improvement plan to a specific project at HS Company, completing a real-world application and analyzing the process and outcomes. The innovation of this study lies in the first systematic introduction of LLMs into project decision quality management, exploring their application value in various decision-making stages. By utilizing LLMs, this paper presents specific application solutions for intelligence gathering, decision communication, interpretability analysis, and execution control, aiming to enhance decision quality and efficiency in project management. The findings not only provide new technological tools for the project management field but also lay a foundation for the development of LLM technology and its applications in other management domains, offering significant theoretical and practical value.
查看全文：	预览  下载(下载需要进行登录)