Simulation methods in entrepreneurship research: a systematic review and text mining analysis

Abstract

Entrepreneurship, a prolific field of research, is vital for nurturing new ventures, yet comprehending its diverse impacts on enterprises poses a challenge. In this context, the use of simulation methods is innovative to unravel the complex organizational processes affected by entrepreneurship. Therefore, this study aims to investigate simulation approaches in entrepreneurship domains. To achieve this, we combined a systematic literature review (SLR) and a text mining analysis based on Natural Language Processing (NLP). In total, 3527 articles published between 1995 and Jan 2024 in the Clarivate Analytics Web of Science database have been collected; 219 articles were selected for the systematic literature review. A machine learning classification method using Latent Dirichlet Allocation (LDA) determines the optimal number of topics. Thereafter, research methods, fields, themes are extracted and categorized. In addition, this paper highlights the major gaps in this field which shed light on the future research directions. Results indicate that the business field is the most prevalent, and Agent-Based Modeling stands out as the most commonly applied technique.