AIBS, in collaboration with the NET ESolutions Corporation, published findings from an analysis of citations of scientific articles which introduced a meta-method for large-scale discovery of scientific communities of practice. The idea of ‘Invisible Colleges,’ self-assembled groups of scientists with common scientific interests, has been around for several hundred years, yet the study of the behavior of such groups has been limited to small samples and case studies. In this work, a new bibliometric method is presented for identifying such groups. This method was validated with expert review of the thematic relatedness of clusters of articles. This study is a first step in designing and testing a meta-method that could enable large-scale identification of communities of different sizes and types, based on different search criteria, data collection technique, choice of clustering algorithm, adjustable parameters, and cluster selection criteria. This method could be used in identifying expertise and/or conflicts of interest for peer review.
The manuscript, entitled “Finding Scientific Communities in Citation Graphs: Articles and Authors” was just published online in the Quantitative Science Studies journal and is open-access and available at https://www.mitpressjournals.org/doi/abs/10.1162/qss_a_00095.