Collaboration model based on entity-concepts connectivity

Collaboration in research and development (R&D) has been increasing rapidly since the 1980s in parallel with accelerating technological change, the internationalisation of firms and globalisation of the world economy. Further, the development of information and communication technology (ICT) infrastructures with new networking features, such as Internet and wireless services, has accentuated the importance of managing knowledge flows rather than the flow of goods. As a consequence, R&D is increasingly undertaken in networks spanning different technology fields, industry and national boundaries. New forms of collaboration are also emerging, ranging from informal bilateral collaboration to highly complex multilateral and multi-layered networks. At the same time competition is intensifying due to the global reach of firms. The knowledge network approach within the framework of the AMI@Work can be used to model community formation based on the competencies of individuals in a collaborative setting. This paper develops an entity-concept connectivity model by including complementary concepts in concept proximity measure of Fuchs, Kittovski and Köhler and by applying the Axelrod landscape modelling to analyse the utility of collaboration. The model suggests some new research directions to further deepen our understanding of knowledge sharing and innovation in the ICT sector. The concept proximity measure is defined as the logarithm of the sum of the relevancies of concept classes of the collaborating parties. The utility of collaboration is defined similarly as the energy measure in landscape modelling. The utility is the sum of the products between the distances of collaboration and the number of alliances between the collaborating parties. One application of the model is the analysis of firm alliances based on concept proximity measure calculated from the patent references. The example shows that there is an inverse-U shaped utility distribution of firm pairs with respect to the concept proximity measure of the corresponding firms.