TIME-BASED ERROR CLASSIFICATION IN VIRTUAL EDUCATION SPACE

Abstract

Training is an incremental process involving slow and methodicalachievement of small goals over a given period. Each step in the process involvesthe assimilation of matter by accepting a set of errors, the correction of which leadsto the improvement of knowledge and deeper understanding of the problem area andmatter as a whole. The classical learning approach, where the facilitator is part of the learning processand has the technical ability to test the work of trained agents, allows for easy andeffective classification of a set of errors to serve as the basis for changing and improvingthe material. One major disadvantage of the e-learning is that the process supervisordoes not have this ability due to the lack of real-time action as well as the large scaleof those training sessions. In this article, we will look at a mechanism for classifying a set of errors usedin the UniPlayground project to categorize the behavior of trained agents based ontheir omissions, errors, or failure to conform to the particulars of the course material.The article examines learning in the context of programming courses and softwaredevelopment.