Automated system for diagnosing the ability to solve computational problems based on structural and mental schemes

Problem and goal. The paper deals with an actual problem - learning to solve computational problems. The components of this problem are highlighted and it is shown that to solve it, it is necessary to automate the learning process, especially in terms of organizing independent work. The purpose of this work is to describe the scientific and technological basis for building an automated diagnostic system for solving computational problems (using the example of physical problems), as well as to present the results of using the developed system in the real pedagogical process. Methodology. The described system is based on a mental approach to learning. The paper introduces the concept of a computational primitive and, based on it, the concept of a structuralmental scheme (SMS) - a graph-like model of the ability to solve problems. An Elo-like rating system was used to provide adaptive, continuous monitoring. In order to increase the strength of mastering the ability to solve problems, we implemented the account of forgetting using a piecewise linear model of forgetting. To assess the formation of the ability to solve problems, a value is introduced - the level of assimilation, which is determined by the structural and mental scheme of the student and reflects its completeness and strength. Results. The results of the application of the described automated diagnostic system for the ability to solve computational problems in the real pedagogical process are presented. The described system proved to be effective, i.e. it positively affects the level of formation of the ability to solve problems. The correlation between the number of tasks solved by the student and the level of assimilation is calculated. The efficiency of the presented system in terms of forming the ability to solve computational problems on the example of physical problems is proved. Conclusion. Automation using a computer system allows to track and store information about the status of each individual SMS connection in memory. This cannot be done using any other non-automated tools and technologies due to the large amount of data and the complexity of calculations.