[Paper Review] Mapping the Landscape of AI-Driven Human Resource Management: A Social Network Analysis of Research Collaboration

Mehrdad Maghsoudi, Motahareh Kamrani Shahri+2 more|arXiv (Cornell University)|Aug 18, 2023

AI and HR TechnologiesBusiness, Management and Accounting3 citations

TL;DR

This study maps the global research landscape of AI-driven human resource management (AI-HRM) using social network analysis (SNA) of co-authorship networks. By analyzing 102,296 authors and 287,799 collaborations, it identifies four core research themes—system identification and control, HR analytics, machine learning for prediction, and AI-driven decision-making—and reveals distinct regional and institutional collaboration communities, highlighting key contributors and emerging trends in AI-HRM research.

ABSTRACT

As artificial intelligence (AI) transforms human resource management (HRM), understanding the research landscape becomes crucial for both academics and practitioners. While existing studies examine isolated aspects of AI in HRM, a comprehensive analysis of collaboration patterns and emerging themes remains lacking. This research employs social network analysis (SNA) to examine the co-authorship network within AI applications in HRM research, providing insights into collaboration dynamics and identifying key research directions. Through analysis of centrality measures and application of the TOPSIS method, the study identifies influential authors, institutions, and emerging research themes. Analysis of 102,296 authors and 287,799 collaborations reveals distinct communities focusing on specific aspects of AI-HRM across regions. The findings identify four primary research themes: AI for System Identification and Control, focusing on workforce planning and adaptive management; HR Analytics and Performance Management, emphasizing data-driven decision making; Machine Learning for Classification and Prediction, addressing talent acquisition and retention; and AI-Driven HR Decision-Making, exploring strategic planning and unbiased evaluation systems. The country co-authorship network analysis uncovers three main communities: Global HR Applications, HRM in the Middle East and Asia, and Global Integration of AI in HRM, reflecting shared regional challenges. Institutional collaboration patterns indicate five distinct communities, from established Asian AI research centers to emerging research hubs in developing economies.

Motivation & Objective

• To map the evolving research landscape of AI-driven human resource management (AI-HRM) through collaboration patterns.
• To identify influential authors, institutions, and emerging research themes in AI-HRM using network analysis.
• To uncover regional and institutional collaboration communities shaping AI-HRM research.
• To provide a comprehensive overview of thematic clusters and their geographic and institutional distribution in AI-HRM literature.
• To support academic and practitioner understanding of AI-HRM research trends through data-driven network mapping.

Proposed method

• Constructed a co-authorship network using 102,296 authors and 287,799 collaborations from AI-HRM research.
• Applied social network analysis (SNA) to compute centrality measures (degree, betweenness, eigenvector) to identify influential authors and institutions.
• Used the TOPSIS method to rank authors and institutions based on multiple centrality criteria.
• Conducted community detection on country-level and institutional-level co-authorship networks to identify collaboration clusters.
• Thematic analysis of publication keywords and content to classify research into four primary clusters: system identification, HR analytics, machine learning for prediction, and AI-driven decision-making.
• Visualized and interpreted network structures to reveal regional and institutional research dynamics in AI-HRM.

Experimental results

Research questions

• RQ1 What are the dominant research themes in AI-driven human resource management based on co-occurrence of keywords and collaboration patterns?
• RQ2 Who are the most influential authors and institutions in AI-HRM research, as measured by network centrality?
• RQ3 How are collaboration networks structured across different regions, and what regional research communities emerge?
• RQ4 What institutional collaboration patterns exist, and how do they reflect global research hubs and emerging research centers?
• RQ5 How do the TOPSIS rankings of authors and institutions compare with traditional centrality measures in identifying key contributors?

Key findings

• Four primary research themes emerged: AI for System Identification and Control (e.g., workforce planning), HR Analytics and Performance Management (data-driven decisions), Machine Learning for Classification and Prediction (talemt acquisition/retention), and AI-Driven HR Decision-Making (strategic planning and fairness).
• The country co-authorship network revealed three main communities: Global HR Applications, HRM in the Middle East and Asia, and Global Integration of AI in HRM, reflecting region-specific research foci.
• Institutional collaboration patterns identified five distinct communities, ranging from established Asian AI research centers to emerging hubs in developing economies.
• Centrality analysis identified key authors and institutions with high betweenness and eigenvector centrality, indicating central roles in knowledge dissemination.
• The TOPSIS method successfully ranked influential authors and institutions by integrating multiple centrality metrics, providing a robust ranking framework.
• The study reveals a growing but geographically uneven collaboration landscape, with strong regional clusters and limited cross-continental integration in AI-HRM research.