Partnerships are essential to achieving the United Nations’ Sustainable Development Goals. In academia, interdisciplinary research can help to address complex challenges related to the Goals. This paper offers a structured approach to identifying current and potential research collaborations across faculties at a Canadian university. Publications from the Dimensions database with an SDG categorization were matched against publications indexed by the university’s Research Information Management System. Potential interdisciplinary research collaborations are then identified by matching authors from different faculties who both have publications within the same research category. Intriguingly, this technique for linking potential collaborators via a shared research category is similar to the hypothesis-discovery model first proposed by Swanson in the 1980s for use in the biomedical field. The utility of this technique for inferring new relationships suggests that it is an archetypal pattern in information science which has applicability in other contexts. Indeed, interest in these techniques is growing as Large Language Models allow causal relationships to be extracted from a broader range of fields.