Sistemnyy podkhod k postroeniyu ontologii dlya avtomatizatsii sostavleniya raspisaniya mnogourovnevogo vuza

The construction of a university class schedule is one of the NP-complete problems. In cases of significant amounts of input data, typical for a multilevel university, and a set of numerous constraints, the search for an acceptable solution may take a long time or may not be optimal. The paper presents the peculiarities of a multilevel university and considers a computerized approach to the construction of an ontological model for the automation of academic scheduling, used to optimize the process of its compilation. The paper utilizes methods of semantic description of the subject area, including computer support for ontological model building. On the basis of the given analysis of the main problems the ontological approach to the formation of data structure for the tasks of training schedules compilation is substantiated. The proposed approach is realized taking into account the conditions of multilevel higher education institution. The ontological model of automated scheduling is developed. The method of solving the problem of scheduling of a multilevel university with the application of genetic algorithm (GA) using penalty functions to take into account the limitations of the mathematical model is presented. The computer program developed on the basis of the constructed class diagram provides the construction of the schedule of academic classes of a multilevel university, effective according to the integral quality criterion.