Investigating Industry–Academia Collaboration in Artificial Intelligence: PDF-Based Bibliometric Analysis from Leading Conferences
Abstract: To investigate trends in industry–academia collaboration within the field of artificial intelligence (AI), we conducted a bibliometric analysis of papers presented at the AAAI and IJCAI conferences between 2010 and 2023. A major challenge in analyzing these papers is that bibliographic metadata, such as author names and affiliations, is not systematically maintained in existing databases. To address this issue, we developed a method for extracting and classifying affiliation strings directly from the text of PDF articles, and subsequently categorized each organization as either academic or industrial to identify collaborative papers. Our analysis reveals substantial changes in the patterns of industry–academia collaboration in AI research over the past decade. In addition, we performed content analysis using document classification to investigate the types of first authors in collaborative papers and to examine differences in research content between collaborative and non-collaborative articles. The affiliation metadata constructed for this study is publicly available on GitHub.
Authors: Kazuhiro Yamauchi, Marie Katsurai
Publication venue: ICADL 2024
Dataset
Reference
The preprint PDF is available at arXiv.
Yamauchi, K., Katsurai, M. (2025). Investigating Industry–Academia Collaboration in Artificial Intelligence: PDF-Based Bibliometric Analysis from Leading Conferences. In: Oliver, G., Frings-Hessami, V., Du, J.T., Tezuka, T. (eds) Sustainability and Empowerment in the Context of Digital Libraries. ICADL 2024. Lecture Notes in Computer Science, vol 15494. Springer, Singapore. https://doi.org/10.1007/978-981-96-0868-3_5