Online Learning Strategies Analysis with Using Process Mining Techniques

To the educational process organizing a learning management system is usually used. This class of IT systems help to control the process educational process at whole. LMS contains information about the learning process, it’s duration and limitations, materials for learning and what is important, students’ profile data and their learning path. Continuous analysis and monitoring is the key to the effective educational process. Educational systems generate a huge amount of data which can be used for current educational processes analysis and evaluation. This work demonstrates an effective approach to student learning behaviour analysis based on accumulated educational process data. It contains the analysis of students learning paths and their ways of interaction with the LMS user interface via using a combination of process mining algorithms for process models synthesizing and binary classifiers from data mining techniques. This work contains a complex long-term educational process case study for a company who offers to obtain supplementary IT education. Students are offered to pass the multi-annual education with a wide variety of courses that is difficult for analysis together. The main problem to be solved is searching the ways to drop out minimization. The factors influenced to drop out probability such as the courses and contracts count, base and specialized course combinations, individual’s paths changes were considered. Additionally, this work explores the approach for user-interface interaction (active actions such as clicks, page transitions) effective estimation based on SLA conclusion to find and estimate transitions and bottlenecks in LMS. For further using of research outcome the dashboards for continuous drop out monitoring and for efficiency LMS using were developed based on Celonis process mining platform. Keywords: educational process mining, educational data mining, drop out prediction, educational process