The Algorithm of Competitive Normalization of Criteria in Rating System of Evaluation of the Intelectual Achievements

Abstract

In the study and analysis of most processes occurring in any field of human activity, there are applied problems that actually belong to the theory of decision making. The most common of these tasks is to build a ranking list of certain objects (subjects), from which you need to choose the best (worst) in terms of the total value of certain attributes that characterize these objects. The complexity of such tasks is that, as a rule, there are no cases when one or more objects have significant advantages over others in all indicators taken into account. That is why there is a need to apply existing methods of decision theory, as well as to develop algorithms that allow mathematical consideration of the specifics of specific practical problems. The article considers the problem of assessing the achievements of secondary schools in intellectual competitions conducted by the Junior Academy of Sciences of Ukraine and the Ministry of Education and Science, and describes the developed algorithm of competitive normalization of criteria for rating participants. This problem was formalized with the help of ontological methodology, which allowed to implement the algorithm for its solution in the mathematical software TNIAS (Transdisciplinary Network-Centric Information-Analytical System). The developed algorithm is based on the competitive nature of the process of establishing the degree of dominance of some alternatives over others, depending on the analysis of numerical characteristics, which were observed over a period of time. In the general case, the use of the algorithm is most successful in problems when the calculation of rating indicators of alternatives depends not on the absolute numerical values of some criteria, but on the number of alternatives that have close values and do not reach or exceed certain thresholds established by analysis subject area.