(IEEE CS, 2009) Ehlers, Kobus; Joubert, Malan; Kinghorn, Johann; Van Zyl, Arnold
Research orientated universities face a conundrum. On the one hand research areas are increasing and expanding; on the other research resources are – in relative terms – diminishing. As a consequence a more active approach to shaping the institutional research profile is called for. This presupposes a clear view of research activities at the institution, but this is notoriously lacking, universities being what they are. This paper reports on a DSS which was developed for research management at Stellenbosch University, South Africa. Within the context of a rather unique national management system for higher education, a data mining approach was developed and tested. A modified ontology model was built to bridge the inherent ambiguity when clustering outputs across different academic disciplines. The results enabled research managers to correlate (in a meaningful way) strategic goals with actual output. The outcome points to a number of further possibilities for the application of this system.