Some results on the graph-theoretical identification of microclusters in the Belgian National Innovation System

In this paper we will present some results of the graph-theoretical analysis of the Belgian network of research co-operation and research agreements as it transpires through the projects falling under the general heading of the R&D Framework programmes of the EU, the projects of the EUREKA initiative, the R&D projects financed by the national authorities at the regional and federal level, and finally the private agreements and other forms of co-operation contained in the MERIT/CATI database. Particular attention will be devoted to the clustering issue: how can we define micro-clusters (2) in a graph-theoretical context, which micro-clusters of actors emerge, and what are their main characteristics. We also briefly address the important question how these micro-clusters relate to clusters identified at higher levels of aggregation and in other types of approaches. A more thorough analysis of this last issue is left for later stages of our research.