科研管理

科研管理 ›› 2024, Vol. 45 ›› Issue (11): 14-25.DOI: DOI：10.19571/j.cnki.1000-2995.2024.11.002

• 论文 • 上一篇    下一篇

人工智能赋能创业：基于结构主题模型的综述

李大元1，潘壮1，陈晓红1,2,3   

  1. 1.中南大学商学院， 湖南 长沙410083；
    2.湖南工商大学前沿交叉学院、管理科学与工程学院，湖南 长沙410205；
    3.湘江实验室，湖南 长沙410205

  • 收稿日期:2023-07-16 修回日期:2024-08-28 出版日期:2024-11-20 发布日期:2024-11-12
  • 通讯作者: 陈晓红
  • 基金资助:
    国家自然科学基金重大项目：“创新驱动创业的重大理论与实践问题研究”（72091310,2021.01—2025.12）课题二“数字创新驱动的新企业创业模式研究”（72091313,2021.01—2025.12）。

AI-enabled entrepreneurship: A review based on the structural topic model

Li Dayuan1, Pan Zhuang1, Chen Xiaohong1,2,3   

  1. 1. School of Business, Central South University, Changsha 410083, Hunan, China;
    2. School of Advanced Interdisciplinary Studies，School of Management Science and Engineering, Hunan University of Technology and Business, Changsha 410205, Hunan, China;
    3. Xiangjiang Laboratory, Changsha 410205, Hunan, China
  • Received:2023-07-16 Revised:2024-08-28 Online:2024-11-20 Published:2024-11-12
  • Contact: Xiaohong Chen

摘要：      人工智能已成为赋能创业发展的重要力量，在创业机会发现与创造、机会资源整合、价值创造等环节中发挥着日益重要的作用。尽管创业领域对人工智能的关注度与研究兴趣日益浓厚，但目前相关研究仍比较零散，缺乏系统整合。本文采用无监督机器学习方法，结合机器编码与人工编码，运用结构主题模型对Web of Science收集的122篇相关文献进行分析，从人工智能赋能创业的前因、情景、过程、结果、理论视角与工具方法等方面梳理了研究现状。研究发现AI赋能创业的动力机制研究有待进一步深化、情景场域研究有待进一步丰富、路径机制研究有待进一步挖掘、双重影响研究有待进一步探讨、底层逻辑有待进一步突破、技术手段有待进一步融合。在此基础上提出了未来研究框架与议题，指出应深化宏观、中观、微观乃至跨层AI赋能创业动力研究，推进AI赋能区域、行业、领域等创业多情景研究，加强跨层次与动态视角下AI赋能创业的过程路径研究，全面关注AI赋能创业可能的积极结果尤其是负面影响，进一步探索基于新实践的AI赋能创业的独特理论体系，为AI赋能创业研究提供坚实的数据、工具及方法支撑。通过对已有文献梳理与未来研究框架构建，本研究为丰富与深化人工智能赋能创业研究提供了参考。

关键词: 人工智能, 机器学习, 结构主题模型, 创业

Abstract: Artificial intelligence has become a crucial force in enabling entrepreneurial development, playing an increasingly important role in the chance of entrepreneurial opportunity discovering and value creating. Although the entrepreneurial field has gradually paid attention to artificial intelligence, the current research on artificial intelligence-enabled entrepreneurship is still fragmented and lacks systematic integration. This paper adopted the unsupervised machine learning methods, combined machine coding with manual coding, and used structure topic models to analyze 122 relevant literatures collected by Web of Science. This study sorted out the current research status from the aspects of the antecedents, scenarios, processes, results, theoretical perspectives, and tools and methods of AI-enabled entrepreneurship. This study found that the research on the driving mechanism of AI-enabled entrepreneurship needs to be further deepened, the research on scenario fields needs to be further enriched, the research on path mechanisms needs to be further explored, the research on dual impacts needs to be further discussed, the underlying logic needs to be further broken through, and the technical instruments need to be further integrated. On this basis, it proposed a future research framework and topics, and revealed that we should deepen the research on the driving force of AI-enabled entrepreneurship at the macro-, meso-, micro- and even cross-level levels, promote the research on the multi-scenario ecology of AIenabled entrepreneurship in regions, industries and fields, strengthen the research on the process path of AI-enabled entrepreneurship from a cross-level and dynamic perspective, pay full attention to the possible positive results of AI-enabled entrepreneurship, especially the negative impact, and further explore the unique theoretical system of AI-enabled entrepreneurship based on new practices, provide solid data, tools and methods to support the research on AI-enabled entrepreneurship. By combing through existing literature and constructing a future research framework, this paper will provide a guideline for enriching and deepening research on artificial intelligence-enabled entrepreneurship.

Key words: artificial intelligence (AI), machine learning, structural topic model, entrepreneurship