The corona epidemic has not only disrupted society but also conventional education. The crisis has highlighted several educational problems such as challenges to design secure grading methods, teacher workloads and student uncertainty. These challenges may be solved by connected and intelligent systems based on Big Data and Machine Learning. However, a main obstacle to develop such systems is to collect high-resolution data on students’ learning patterns. The digital learning platform Yata makes it possible to collect student data without being invasive on the learning process. We therefore propose to use the data collection from Yata to study the combination of Big Data, Machine Learning and education. In particular, we plan to use correlation and cluster analysis to study the learning patterns in the data set and compare different Machine Learning models to investigate the inference of predicting student performance (e.g. grades). By understanding the data set and models, educational applications may be designed to support teachers and students and to develop a more effective education.
The project is closed: 30/11/2020