Academic-Industry Engagement: The Role of Machine Learning in Predicting Contract Research Outcomes

Johnson, David; Ao, Jingning; Bock, Adam J.; Schlegel, Viktor

Abstract

Academic-industry engagement, such as contract research, facilitates the development of university-centered entrepreneurial ecosystems (UCEEs). Research implies that the language utilized within contract research proposals is critical in determining whether an academic chooses to engage with an industrial partner or not. However, we know very little about the role of contract research proposal narratives in facilitating successful academic-industry engagement outcomes. Accordingly, adopting an explorative study, we apply machine learning (ML) techniques to predict successful academic-industry contract research outcomes and reveal key linguistic features associated with successful contract research proposals. Our predictive and exploratory ML techniques achieve an 83% accuracy in predicting successful academic-industry contract research outcomes and reveal that the use of concise and field-specific vocabulary repetitively is associated with successful contract research proposals. Our findings develop research and policy relating to academic-industry engagement. At the same time, our ML techniques provide a useful foundation for scholars to further develop theory, practice, and policy within the academic entrepreneurship and entrepreneurial ecosystem fields.

Citation

Johnson, D., Ao, J., Bock, A. J., & Schlegel, V. (2024). Academic-Industry Engagement: The Role of Machine Learning in Predicting Contract Research Outcomes. Academy of Management Proceedings, 2024(1), https://doi.org/10.5465/amproc.2024.10620abstract

• Journal Article Type: Article
• Acceptance Date: Mar 29, 2024
• Online Publication Date: Jul 9, 2024
• Publication Date: 2024-08
• Deposit Date: Jul 15, 2024
• Journal: Academy of Management Proceedings
• Print ISSN: 0065-0668
• Electronic ISSN: 2151-6561
• Publisher: Academy of Management
• Peer Reviewed: Peer Reviewed
• Volume: 2024
• Issue: 1
• DOI: https://doi.org/10.5465/amproc.2024.10620abstract
• Public URL: https://durham-repository.worktribe.com/output/2583888