The transition to digital society is characterised by the development of new methods and tools for big data processing. New technologies have a substantial impact on the education sector. The article represents the results of applying big data to analyse and transform the learning content of Moscow’s schools. The analysis of the school curriculum comprised the following: (a) identifying one-topic lesson scripts, (b) analysing cross-disciplinary connections between subjects, (c) verifying the compliance of the lesson script digital content to the Federal Educational Standards. The analysed material included 36,644 lesson scripts. The analysis has been conducted using specifically designed digital tools featuring data mining algorithms. The article considers the issue of applying data mining algorithms to analyse school curriculum for the improvement of its quality.
This paper reviews the key research of the automatic engagement detection in education. Automatic engagement detection is necessary in enhancing educational process, there is a lack of out-of-the-box technical solutions. Engagement can be detected while tracing learning-centered affects: interest, confusion, frustration, delight, anger, boredom, and their facial and bodily expressions. Most of the researchers reveal these emotions on video using Facial Action Coding System (FACS). But there doesn’t exist a set of ready-made criteria to detect engagement and many scientists use additional techniques like self-reports, audio-data, physiological indicators and others. In this paper we present a review of most recent researches in the field of automatic affect and engagement detection and present our theoretical model of engagement in educational process based on the learning-centered affects’s detection. Engagement is understood as an affective and cognitive state, accompanying learning process. While reaching optimal engagement students experience various affects, where highly positive and negative feelings mean that a student is close to be engaged in the learning process.
Problem and goal. Developed and tested solutions for building individual educational trajectories of students, focused on improving the educational process by forming a personalized set of recommendations from the optional disciplines. Methodology. Data mining and machine learning methods were used to process both numeric and textual data. The approaches based on collaborative and content filtering to generate recommendations for students were also used. Results. Testing of the developed system was carried out in the context of several periods of elective courses selection, in which 4,769 first- and second-year students took part. A set of recommendations was automatically generated for each student, and then the quality of the recommendations was evaluated based on the percentage of students who used these recommendations. According to the results of testing, the recommendations were used by 1,976 students, which was 41.43% of the total number of participants. Conclusion. In the study, a recommendation system was developed that performs automatic ranking of subjects of choice and forms a personalized set of recommendations for each student based on their interests for building individual educational trajectories.
Predicting the educational success of students is one of the actual tasks of the intellectual analysis of educational data. In this article, two research issues are considered: improving the quality of the university students’ academic performance prediction model and implementation the developed model into the real university educational process. The models predicting academic performance are based on XGBoost algorithm and the linear regression algorithm. According to the results of the study, it was revealed that data on the use of electronic and university libraries make it possible to improve the quality of predicting the students’ academic performance, and also confirm the fact that monitoring the students’ academic performance in dynamics is more informative in making managerial decisions in the educational process than the absolute values of the academic performance results. The models for predicting the students’ academic performance studied in this work can be used in educational institutions of higher education for the timely identification of at-risk students, providing feedback to students and teachers regarding the educational success of students and managing the educational process.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.