ABSTRACT
Contemporary innovation processes increasingly involve a large number of networked actors, and cross-fertilization between knowledge institutions and firms has thus become a significant driver for innovation. Important insights into the differing nature of research and development (R&D) collaboration in particular sectors have been provided by research inspired by the knowledge-base approach embedded within innovation system (IS) theory. This study aims to contribute to this body of literature by applying the concept of differentiated knowledge bases to the former state-socialist countries, where the IS operates through a firewall between academia and industry. Data on collaborative R&D projects co-financed by public resources have allowed a detailed analysis of the nature of collaboration networks, revealing emerging patterns of academia–industry linkages and questioning the propositions stemming from the knowledge-based approach. The study concludes that collaborative science–industry networks show a very distinct topography when analytical and synthetic knowledge is compared.
Disclosure statement
No potential conflict of interest was reported by the authors.
ORCID
David Marek http://orcid.org/0000-0002-0441-0942
Notes
3. A few thousand results in the case of the most productive entities.
4. Discipline-based standardized citation rate 2008–2012 on Thomson Reuters Web of Science.
5. In logistical regression the influence of physical distance between two entities on formation of their partnership was tested against the total number of results as proxy to quality of knowledge supply.
6. Especially entities with activities concentrated in the narrow spectrum of research branches connect primarily to partners in related fields.
7. The ForceAtlas layout emphasizes complementarities, it is made to spatialize small-world networks and is focused on quality to allow a rigorous interpretation of the graph.
8. As opposed to firm–firm or academia–academia interactions.
9. In terms of frequency of connection as well as total costs of collaborative tasks.
10. We use assignment of results into research branches to describe precisely the proportion of each KB for respective entity (see methodological section). For entities without result listed in the database, we are unable to perform an assignment.
11. We decided on an arbitrary cut-off point based on histogram; the share of prevailing KB has to be 60% or higher to accentuate underlying differences in knowledge-sourcing.
12. The degree of a node is the number of edges incident to this node and can be understood as proxy for network density.
13. The average number of steps along the shortest paths for all possible pairs of nodes can be understood as a proxy for the ability of the network to transfer knowledge.
14. Ratio of the number of existing and possible edges D = T/N (N–1).