Leveraging machine learning for bibliometric analysis of emerging fields

Bibliometric analyses of emerging fields with inconsistent terminology and porous boundaries are challenging: When precise terms for search queries are not available, compiling a comprehensive dataset requires screening a large number of database records to prevent false positives. In this study, we leverage Machine Learning (ML) to identify and include publications that are relevant to the field but differ in their terminology. ML is employed to semi-automate the necessary screening process of the emerging research landscape of translational psychotherapy as a use case. Compared to a typical database search with terms of known terminology only, the dataset generated by the ML-augmented approach differs clearly in various bibliometrically relevant aspects, such as top authors, journals, countries and impact. Our study emphasizes the importance of consistent terminology of research fields and, in its absence, the merits and benefits of ML. notReviewed other