Toward a network theory of innovation:Heterogeneity in relations, positions, and perspectives
Knowledge-based innovations span networks across functional domains such as novelty production in R&D and wealth generation on the market. The indicators of these domains are heavily institutionalized. For example, maps of patents in USPTO cannot easily be compared with publications in the Web-of-Science; innovations require the mapping of interfaces: different perspectives have reflexively to be recombined. When the systems are animated in parallel (for example, on split screens) one can show delays and feedbacks among the different domains.}
In early stages of a knowledge-driven technology, for example, researchers may preferentially attach to the inventors, whereas in a next stage, preferential attachment moves to global “centers of excellence” such as Boston, London, and Seoul. The economic dynamic may be orthogonal; for example, in terms of marketable applications.
Using CuInSe2 is a material used for the coating of Photovoltaic Cells in thin layers. We study these technologies for Alkemade et al. (in preparation) in terms of both patent classes (cognitive diffusion) and inventor addresses (geographical diffusion). The development of inventor addresses in USPTO data is shown at http://www.leydesdorff.net/photovoltaic/cuinse2 (or similarly for PatStat data at http://www.leydesdorff.net/photovoltaic/cuinse2.patstat ). One can see the animations at http://www.leydesdorff.net/photovoltaic/cuinse2/animate.html.
The development of International Patent Classifications in this same set is visualized at http://www.leydesdorff.net/photovoltaic/cuinse2/cuinse2.ppsx. The development of the (Rao-Stirling) diversity shows the three generations of the technology that can inform the interpretation of the geographical diffusion. (Figure 1)