In this paper we apply co-author analysis
to create from a large set of publications clusters of collaborating researchers within a faculty of chemical engineering. Results have been discussed with an expert. The co-author clusters appeared to be meaningful, with respect to the identification of research groups, the relations within these groups, as well as to relations between these groups and changes in time. Also differences between ISI-based and non-ISI based maps proved to be consistent with the expert's opinion. Many clusters represent collaborating authors grouped around a full professor, mostly the department chairman. Co-author analysis can be used, for example, as an important tool in evaluative bibliometrics in order to make a first identification of research groups in
unknown
universities or organizations.