Test models and algorithms for knowledge presented as Petri networks

Due to the development of computer engineering and data processing technology, it has become possible to create computerized training systems based on artificial intelligence. One of problems that can be solved in the computerized training systems by means of Petri networks is testing the knowledge base in case test tasks are of the open type and the form of the trainee answer is maximally free. The paper represents solutions of the problems of computerized natural language interpretation and assessment of the trainees' answers.