The identification of one's own research specializations is of crucial importance
for research administration at universities. In this case study, two different approaches
were applied to the University of Vienna. The first relies on third-party funding
and ranking results as well as on other esteem indicators. The second is based on
a bibliometric analysis of the publication output. We used the Web of Science subject
categories for publications clustering. In addition publications were also clustered
for two esteem indicator-based research specializations (one STEM, one SSH) by means
of the lists of the researchers associated with the research specializations (RS).Both
approaches proved to be useful in identifying research specializations, lead to similar
results and are meant to be used in a complementary way. We found that the greatest
hindrance lies in the inherent limitations of journals-assignment-based classification
systems that does not correspond to university structure and the considerable time
and efforts for more accurate researcher-based publication analyses. Further investigation
on this subject, including new and alternative metrics, is needed and will be conducted
in the future. However, the preliminary results already demonstrate the benefits of
bibliometric analyses for the identification of research specializations of universities.
