Identifying interdisciplinary research in research projects

Identifying interdisciplinary research has become an important area of study in scientometrics. However, defining what exactly constitutes interdisciplinarity and how it manifests in research activities, such as publications or research projects, remains challenging. In this paper, we propose a mathematical modeling approach to interdisciplinarity measurement based on assessing project diversity. Particularly, we propose a novel approach that combines three indicators: the diversity of researchers, the diversity of research organizations, and the diversity of research disciplines involved in the project, to identify potentially interdisciplinary research projects. To measure diversity, we employ various methods, including distance matrix calculation, evaluation of the distance between researchers, and assessment of the relevancy of researchers' expertise to the projects. We implemented the proposed approach on two datasets; FRIS and Dimensions. We could classify the interdisciplinarity of projects into three groups-Low, Medium, and High. Empirical results analysis supports the proposed approach assumption that the diversity of research projects gets higher when the distances between disciplines in the projects increase. Further, it was shown that the diversity of researchers and organizations was strongly affected by the distance. The number of researchers and organizations had a relatively small impact on the overall diversity score. Furthermore, the relevancy weight can be incorporated as an additional factor in the measurement of interdisciplinary.