摘要
在数字化时代,数据已经成为一种新的生产要素,被认为是第四次技术革命的关键资源。大数据的开放、共享和流通成为万物互联的基本需求,而大数据多样性的基本特征,也使人思考在管理学研究当中需要解决的基本理论与应用问题。本文深入梳理与分析了“大数据”概念被提出以来,人们认识世界角度与方式发生的颠覆式转变,以及大数据科学与AI技术在应用中所产生的管理学问题,总结指出了大数据的科学描述、数据的联动协调、全域管理目标场景谱系、大数据共享的管理机制,以及AI技术与算法创新等关键问题对管理学在研究范式层面的挑战。面对未来发展问题与挑战,本文提出了相应理论、管理机制等方法对策,并结合相关案例研究说明其未来应用的可行性。
In the digital age,data has become a new factor of production and is considered a key resource of the fourth technological revolution.The openness,sharing,and circulation of big data have become the basic needs for the interconnection of all things.How to excavate the intrinsic connections between seemingly unrelated and even contradictory things through the horizontal full-domain linkage of data across industries,fields,time and space,and even ecosystems,and to make them produce chemical reactions,is an important driving force for data to empower corporate innovation.Therefore,the issue of the full-domain linkage of data resources brings many challenges to the study of management in the digital age.The vertical linkage and completeness of big data have relatively mature technical means.Given an industry,each vertical association has a relatively mature technical solution.However,under the conditions of deep digitalization across industries,the horizontal completeness of data between enterprises and even industries still has many problems,such as incomplete horizontal completeness,defects in enterprise data sharing mechanisms,and the rapid iteration of artificial intelligence ecosystems posing challenges to the development of industrial policies.In response to these issues,this paper proposes improvements to the research paradigm of management studies from the perspective of changes in management research.First,the horizontal linkage of data is consistent with the human brain`s learning process through extensive interaction with multi-channel information,thereby achieving high-level recognition and understanding of cognitive objects.Based on this,granular computing ideas and generalized set theory are organically combined to propose a quantifiable linkage space for horizontal linkage of data across scenarios.Second,for target management,drawing on the prism model in physics,a concept model of target spectrum is proposed,viewing the process of knowledge formation as a process from recognizing“visible light”to recognizing“invisible light,”striving to provide a quantifiable and reasonable explanation for complex and vague management objectives and influencing mechanisms.Third,in response to the shortcomings of blockchain technology in management practice,starting from the typical carriers of managing massive data in management,a“centralized system data non-centralized sharing”mechanism is proposed,further clarifying the rights attributes of data to serve management empirical research.Based on this,an application plan for this sharing mechanism and multi-modal large models in management research is given.Fourth,the proposal of the horizontal linkage technical mechanism of data has completely freed the study of corporate ecosystems from the limitations of hypothesis-driven in traditional research paradigms.For the problems of parameter selection,cross-scenario reproducibility,and strong assumptions in traditional empirical studies,this paper proposes a two-way interactive double gyroscope model between policymakers,managers,and cluster decision-making.Furthermore,due to the complexity of data linkage after digitization,the mode of AI methods in corporate management research and policy research needs innovation.This paper organically combines generative adversarial networks with deep reinforcement learning models,proposing a methodology for the study of corporate ecosystems under the background of data intelligence,and summarizes a methodological framework for innovation in corporate management research patterns.Finally,guided by the concept of full-domain linkage,this paper provides different research directions and application prospects,including the continued improvement of horizontal linkage technical mechanisms,data management order,large model systems for management empirical research,and full-domain scenario target spectra.
作者
黄楠
李竹伊
李雪岩
宫大庆
Huang Nan;Li Zhuyi;Li Xueyan;Gong Daqing(School of Management,Minzu University of China;School of Business,Renmin University of China;School of Management,Beijing Union University;School of Economics and Management,Beijing Jiaotong University)
出处
《南开管理评论》
CSSCI
北大核心
2024年第7期74-85,共12页
Nankai Business Review
基金
国家自然科学基金项目(72103019、62276020)
中国国家铁路集团重大项目(N2021S017)资助。
关键词
大数据科学
大数据管理
大数据全域联动
挑战与对策
Big Data Science
Big Data Management
Big Data Global Linkage
Challenges A、and Countermeasures
