Management of academic staff activity: modeling and prediction of rating system indicators

This paper deals with the problem of constructing a system of rating indicators for stimulating the work of the academic staff in higher educational institution. Many areas of teacher activity (for example, educational, scientific, international, etc.) laid the basis of selection the groups of indicators in the system. Social challenge in improving the quality of educational services determines the relevance of research in the field of modeling and prediction of indicators which characterize the work of high school teacher. To predict the dynamics of the structure of the rating indicators in the system, the authors introduced the concept of drift and variability of each group. Using informational hypercube for the structure of input data allowed authors to take into account the individual characteristics of each parameter included in a mathematical model to describe the rating indicators. To make the prediction of the structure and values of rating system indicators the authors introduced the concept of drift. Drift of indicators takes into account the introduction of new indicators, the removal of existing indicators, and movement of indicators between the groups. In the article, authors introduced a novel quantitative indicator of group variability. The value of this indicator determines the prediction strategy of the teacher work in higher school in the future period. To predict the total amount of stimulating, the complex technique offered and it includes four modules: modeling values within the existing range in the previous period; modeling new index value based on the assumptions introduced using a random number generator; exclusion a range of values of deleted indicators; modeling new values based on the study of the modern trend of indicators. The presence of flexible information structure in the form of a hypercube and complex mathematical model allowed authors to carry out numerical simulation for predicting the values of individual and group indicators. During the experiment, the structural stability of values is detected, and it does not lead to a drastic changing of the quantitative ratio between the groups of indicators.