Topic modelling algorithms can be applied to a large collection of documents to uncover themes and their relationship, thus adding an automatic semantic layer to the archive of documents. In this work, topic modelling has been applied to the collection of papers published at the EUNIS (European University Information Systems) annual congresses from 1997 until 2015. Drawing on the initial work of (Wilson and Bergström, 2015), this paper extends the historical analysis of the conference proceedings with the application of text mining techniques to analyse the original texts. The performed analyses include the automatic identification of the most relevant topics and their evolution throughout the years for European University Information Systems, and the quantification of the degree of internationalization and cooperation of the EUNIS annual conference.

--