Detecting Central Research Results in Research Alliances Through Text Mining on Publications

Abstract

Not only in the German research landscape, the establishment of research alliances has become a key element of the national funding structure, especially in order to address current societal, economic and scientific problems. These complex problems are mutually investigated by heterogeneous actors whose heterogeneity can be mainly seen in a combined research effort of scientific as well as business-driven research – a so-called transdisciplinary research approach. The main challenge which arises from this approach covers the cooperation of numerous actors in a complex and often intransparent collaboration structure. To allow transparence of the central research topics within these structures, publication data has to be consolidated and classified.

In order to address this challenge, the establishment of an information management environment supports the ability to handle big repositories of publication data on the one hand and to visualize different thematic interests on the other hand. In this example, fostering cooperation among actors, by revealing thematic accordance, connections and development, becomes possible.

The paper addresses the question in how far an information management environment can support this revealing process by means of classification publication data. Focusing on an information management environment in its pre-prototypic stage, the development process as well as initial results are presented. The results are derived from publication data examined by the transdisciplinary research alliance “Innovative capability in demographic change” initiated by the German Federal Ministry of Education and Research (BMBF).