Artificial Intelligence Research in Management: A Computational Literature Review
Are you feeling buried under thousands of AI-related papers, unable to discern what is essential to your work? This paper is here to guide you through the noise and present findings from 6324 AI academic papers in a concise manner.
This study provides a comprehensive analysis for AI research in management and presents a computational literature review with an abstract-based sampling approach to investigate the status of the management literature to take stock of academic research of the past two decades.
Findings
We present the evolution of research pre- and post-AI spring, emerging topics as well as saturated areas with 41 distinct topics. The findings show that the previously disjointed topic network structure is fully connected by early 2010s and the upward trend in management research starts in the period of 2014-2015.
Here’s an overview of hot and cold topics in today’s AI research in management. 
While saturated areas can guide practitioners to adopt tools and techniques from areas where an academic consensus has been reached, nascent areas might provide opportunities for exploration.
We invite you to refer to our topic modeling analysis to understand what each domain covers and how they link to each other. While saturated areas can guide practitioners to adopt tools and techniques from areas where an academic consensus has been reached, nascent areas might provide opportunities for exploration.
Methodology
We analyze 6324 papers from 1990-2020 published in five management-related domains and identify 41 distinct topics.
Applications and beneficiaries
Practitioners from industries under the Application category can identify relevant technical topics via most recent industry-technique co-occurrence. For instance, Service (18) is quite strongly connected to Scheduling (24), but also with Social Media (9), which are connected strongly with NLP (16). These can support managerial decision making about choosing technical approaches and searching dedicated examples of their applications. In the case of industries absent from the findings, practitioners may either refer to adjacent industries or start from the technical side. For the latter approach, they can either look at the core of the network to identify the most connected applications, or screen for similar applications and identify connected technical topics.
Reference to the research
J. Arsenyan and A. Piepenbrink, “Artificial Intelligence Research in Management: A Computational Literature Review,” in IEEE Transactions on Engineering Management, vol. 71, pp. 5088-5100, 2024, doi: 10.1109/TEM.2022.3229821.
