The Study of Scientific Projects Evaluation Criteria by Means of Machine Learning Methods Through the Examples of Russian Foundation for Basic Research Grant Competitions

Authors

Grigoriev O. Devyatkin D.

Annotation

The paper considers the result of analysis of the main criteria for research projects evaluation by scientific foundations. A brief review of the experts' surveys analysis methods in making decision on projects funding is given. The novel approach to the determination of importance of scientific projects evaluation criteria is proposed; it is based on machine learning techniques. The approach allows scientist to apply the qualitative and quantitative criteria in case of large volume of data and does not require bringing the criteria to the numerical value. The authors experimentally determined the relevance of scientific criteria for projects estimation through the examples of Russian Foundation for Basic Research grant competitions. It has been established that the set of the most important criteria for projects evaluation remains virtually unchanged for all scientific fields besides the “Natural science methods in humanitarian sciences” area. The authors concluded that the suggested approach can be applied for the verification of experts’ final estimate as well as for the check of the importance of the newly introduced criteria.

External links

DOI: https://doi.org/10.22204/2410-4639-2016-092-04-135-146

Russian Foundation for Basic Research Journal PDF at the official website (in Russian): https://www.rfbr.ru/rffi/pdf_read/?objectId=1967517#page=136

Reference link

D. A. Devyatkin, R. E. Suvorov, I. A. Tikhomirov, O. G. Grigoriev. The Study of Scientific Projects Evaluation Criteria by Means of Machine Learning Methods Through the Examples of Russian Foundation for Basic Research Grant Competitions // RFBR Journal, № 4, October-December 2016. Pp. 135-143