Assessment of Online Courses Control Materials Using Information Theory Methods

The digital transformation of industry poses new challenges for related industrial fields, in particular for education. Universities must transform accordingly and provide the digital economy with qualified personal. Online technologies, including online learning, play an important role in digital transformation, opening up new opportunities and posing challenges that the education system has never faced before. The problem of making justified decisions on the selection and assessment of the quality of online courses and their control materials is one of these challenges. Data mining and machine learning analytics tools that work with learners’ digital footprint data can be used to solve this problem. To do so, this work uses the methods of information theory. The indicator of the informativeness of control materials provides differentiation of standalone checkpoints and online courses in general and minimizes the subjective factor present when using existing assessment methods. Another indicator that assesses the mutual informativeness of checkpoints allows determining how much information about the final test results can be obtained by monitoring current academic performance. This indicator is applicable both to individual checkpoints and their series (assembled by type—test assignments, homework assignments, etc.; or by the chronology of the course—¼, ½, etc. of the course length). With its help, it is possible to evaluate the course checkpoints and make a decision on how to improve the course materials. In addition, this indicator can be useful for assessing the marginal accuracy of the prediction of the test results according to the current academic performance of a student. The proposed methods have been tested on online courses of the Ural Federal University with positive approbation results.