Natural language processing in systems with computer-aided knowledge testing

In the sphere of distance learning, computer-aided knowledge testing (CAKT) is widely used. The state of affairs in the field of CAKT systems is analyzed in this paper. It is shown that widespread "elective" tests have a number of essential disadvantages. The authors offer an approach permitting one to use natural language processing as one of the major CAKT components. The main idea of the approach consists in high-tech matching of a student\´s answer with the answer template given by the instructor. In addition, text information obtained from electronic textbooks and books on the course and transformed into a semantic network can be used when evaluating students\´ work. Necessary methods and algorithms were designed. The authors also offer software implementing the methods developed. This software can be built as a subsystem into electronic textbooks, electronic courses and other CAKT systems. The results obtained allow one to believe in appropriateness of the approach for actual usage