The Study on Knowledge Self-Organization and Technological Structure of the Russian Regions by Means of Kohonen’s Self-Organizing Maps

The article proposes a novel model for assessing the quality of technological development which differs from the similar spillovers by introducing a fundamentally new parameter of knowledge selforganization. Unlike spillovers measuring financial R&D flows, knowledge spillover measures structural similarities presented in patents, articles and other quantized units. Being the results of the reactions on the emergence of technological tasks and absorbing new technologies themselves, patents reflect real industrial picture of distribution of new technologies in any particular area. Implementation of selforganizing neural maps unveiled strong self-organized structural patterns distributed across the Russian Federation which were undetectable by means of conventional spatial econometric methods. Furthermore, neural maps exposed serious drawbacks of the Russian knowledge flow system, which is a drastic lack of flow in several high tech areas such as biotechnology. Self-organization indicator can be applied for evaluation of Megascience projects or other programs on both regional and federal levels.   The structure of regional technologies based on 24 technological areas is studied and mapped on neural model, thereby it has been hypothesized that self-organization has an effect on qualitative processes of technological development. The study presents validation model of this hypothesis based on Kohonen’s self-organizing maps. Enhancement of this model on the further spatial studies is shown. Knowledge self-organization variable is developed to indicate technology integration and emergence.