Adaptive distance e-learning systems: how MAI prepares future education basis

24 September 2020

Adaptive distance e-learning systems: how MAI prepares future education basis

Distance learning nowadays gains popularity, and it has become especially relevant in the context of the current COVID-19 pandemic. The Moscow Aviation Institute has had a platform for remote training of students for a long time already — the LMS MAI ("learning management system"), which is built on the basis of the public web Moodle application, which allows to create online classes.

The unique part of the LMS MAI is the adaptive distance learning system Class.Net. It analyzes the user’s level of knowledge and allows the student to make the e-learning process as effective as possible, building an individual educational trajectory. In addition, starting from this year, MAI has launched the course for master degree students, which allows them to understand all the peculiarities of the system of this kind. The author of the course and one of technical developers of the Class.Net Deputy Head of the Department 804 "Probability theory and computer modelling", Professor, Head of the laboratory of distance learning Andrey Naumov shared the details.

How the adaptive Class.Net system works

Every year about 4000 students of all the MAI technical institutes get educated via the system.

Initially, it was conceived as a tool to complement the face-to-face education process. It allows students to complete their homework, prepare for control activities, conduct self-testing, and it can also be used as a reference. It also gives teachers the opportunity to create tasks for students using the methodological content base, to get statistical information about the work of each student and about the tasks that caused the greatest difficulties. Thus, teachers can adjust their approach to classes and change the standard plan, considering it in details, especially those tasks that are difficult for students, and improve their academic performance.

— Normally, distance learning should go in parallel with the fulltime course and serve to organize independent work of students. — emphasizes Andrey Naumov.

The system uses a number of mathematical methods for statistical information processing, such as Item Response Theory (theory of test tasks), which allows to form the current rating of the users, evaluate their training and the level of complexity of the proposed tasks, and eliminate the yet unaffordable ones from the course.

When forming an individual trajectory, machine learning elements are used — various classifiers, in particular. Based the statistical information analysis, it is possible to divide users into several categories and create an individual task of a certain level of complexity for each of them. This is how the system is adapted at each stage of training. A special feature of the system is that each task has its own parameters generated randomly. In other words, it is impossible to copy the answer from a classmate without knowing the solution.

At the same time, if necessary, the system might be used as a self-sufficient training tool, including for theoretical material presentation purposes, tests, control tasks, and tasks with metered pedagogical assistance (tips). The algorithm works via checking the authenticity of responses. For example, to make sure that a student does the work personally, a probabilistic model of the system user’s response time to tasks is used. If responses start arriving too quickly, the probability of third-party assistance or external resources is analyzed using the statistical hypothesis testing tool.

Application scope

The possibilities of the distance learning systems application are numerous. Systems might be used in schools and universities, for professional development, for training and enterprise personnel instruction, and so on. At the same time, the introduction of the adaptive models may significantly improve the quality of the distance learning.

— Our system was developed for mathematical subjects teaching to the students of MAI engineering programs, but it is now already being used in other areas as well, — says Andrey Naumov. — It contains electronic textbooks, and their content is structured in a special way, provided with hyperlinks and divided into levels of complexity based on a graph-oriented approach to make it possible to build an individual educational trajectory. Department 804 already has two theses defended on adaptive distance learning systems theme, one of which concerns the principles of electronic textbooks design.

MAI is ready to provide services for the use of Class.Net to all the interested organizations. We are experienced already. For example, the system was successfully used for students of the Ufa State Aviation Technical University (UGATU).

The possibility of collaboration is simplified by the fact that the system is based on cloud technologies. They have obvious advantage over the installation on the customer’s server, which is due to constant updates, the need to train maintenance personnel on the part of the consumer, and so on. All updates are automatically available to the user from the MAI server, and access can be obtained from anywhere in the country.

Expert education in adaptive systems

Andrey Naumov has developed two education courses on the MAI system.

— The first course was for teachers in the framework of professional development, — he says. — For the past 10 years, I have been teaching this course for the teaching staff of all MAI institutes. After all, for the system to function, it must be constantly "fueled" by electronic textbooks. As a result, we have created about seven courses specifically for Class.Net.

Since 2020, the MAI IT center has launched the master’s program "Computer modelling and optimization of information systems". Based on previous experience, the course "Technologies for developing adaptive distance learning systems" was developed specifically for second-year students of the program.

— At the course for master’s students, I talk not only about the experience of online education technologies application, but also about the design of adaptive distance learning systems, methods of their mathematical support, didactic aspects, cloud technologies, approaches to staff training, and so on. — the author explains.

Thus, graduates of the master’s program are experts in the field of distance learning systems. They are well-versed in the use of cloud technologies and special software tools for such systems, understand their architecture and mathematics, which allows them to adapt the system to the user’s level of knowledge. In the future, they can work in the field of providing resources based on cloud technologies, administer distance learning systems, design electronic control shells for them, and also participate in the development of content together with teachers, understanding what it should be like.

Ask any question