From Digital Footprints to New Tools for Education

TSU specialists worked on the creation of a methodology and the basis of a new technology in cooperation with the University Consortium of Big Data Researchers. The study began back in 2020, when the COVID-19 pandemic drastically changed the attitude towards the organization of the educational process around the world, and higher education faced a number of extraordinary challenges that had had no such precedents before. Then, for the higher education system, the need to assess student satisfaction with the quality of education and the effectiveness of universities in force majeure conditions came to the fore.

“To solve this problem, we decided to use the analysis of social networks, blogs, and forums, as the production of user-generated content is constantly growing there. Some of the available information can be used to detect general trends, understand the scale of the crisis, or find out typical and specific changes in users’ moods,” says Anna Bogdanova, Head of the Online Education Technologies department at TSU.

The researchers analyzed the digital footprints of students from the VKontakte social network, the leader in Russia in terms of the number of registered users and published messages, using certain Big Data tools on the PolyAnalyst software platform. The original data were more than 2 million messages from 548 communities of Russian universities.

“In fact, we have developed the basis of the technology that allows us to identify problematic issues, including the time of their occurrence and relevance, as well as the degree of concern of network users, in our case, students. The methodology is universal and suitable not only for assessing students' attitudes towards the quality of the educational process or its support but also for assessing the reaction of any social group to any burst of information, including local problems and global crises,” emphasizes Professor Mikhail Krishtal, Doctor of Physical and Mathematical Sciences. “So far, we have tested and verified the technology in semi-automatic mode. But we also justified the possibility of switching it to a fully automatic mode. It will allow, upon request, to make the necessary selection and analysis of data simultaneously with their generation, i.e., to obtain the necessary information from the generic data stream at the rate of its appearance. Such content analysis is useful for tracking the occurrence of any problems or, on the contrary, good news, that is, to identify all sorts of reactions in the student community as well as in other groups. With such data, it is possible to respond promptly to problematic situations and make predictions.”

Working with the digital footprint is an important component of online education, unlike the traditional approach. TSU is convinced that tracking student activity, identifying typical scenarios of educational behavior, forecasting, testing hypotheses, and creating an adaptive learning system on a scientific basis, as well as continuous improvement based on feedback, make online education more productive in comparison with traditional methods. This approach becomes a competitive advantage for the university, helping it to strengthen its position in the educational services market compared to universities that do not use big data technologies, digital footprint collection and analysis, and achieve the quality of online learning comparable to the quality of full-time education.

Scientists believe that the next step in the development of the topic should be the creation of a fully automated technology.

“We have shown that, with fairly high accuracy (the measurement error is about 15%), it is possible to draw conclusions about the reaction of the online community to a significant event. User-generated data is an important and easily available source of public opinion and can successfully replace our usual sociological surveys. At the stage of testing the methodology, work is done with data generated in the past. Having learned how to analyze them in real time, we will get a powerful tool for measuring tension through indexes and metrics and using this information to track spikes in tension and premature reactions,” concludes Mikhail Krishtal.

The results of the study are reflected in an article published in the journal Higher Education in Russia. The monthly scientific and pedagogical journal publishes the results of fundamental, exploratory, and applied problem-oriented research. The journal discusses topical issues of theory and practice in modernizing domestic and foreign higher education. The journal is published simultaneously in Russian and English. It is carefully read all over the world, as evidenced by the fact that it is represented in international databases of scientific periodicals, including Scopus, in the highest first quartile.

The University Consortium of Big Data Researchers is a union of educational institutions and researchers engaged in both fundamental research and practical applications of big data collection and analysis. The main goal of the consortium is to jointly conduct scientific and applied research as well as solve problems of public importance using data. Today, it consists of 70 educational and scientific organizations.