# Tracking Behavioral Patterns among Students in an Online Educational System

**Authors:** Stephan Lorenzen, Niklas Hjuler, Stephen Alstrup  
**arXiv:** [1908.08937](https://arxiv.org/abs/1908.08937)  
**Date:** 2019-08-21  
**Summary page:** https://www.summarizepaper.com/en/arxiv-id/1908.08937v1/

## Key points

- Authors conducted a study titled "Tracking Behavioral Patterns among Students in an Online Educational System"
- Focused on analyzing log data from Clio Online, the largest provider of digital learning content for primary schools in Denmark
- Dataset included information from 14,810 students and 3 million sessions spanning from 2015 to 2017
- Used non-negative matrix factorization techniques to uncover dependencies among factors such as time of day, subject, activity type, complexity, and performance
- Longitudinal approach allowed for tracking behavioral changes over time for individual students and within the entire educational system
- Findings provided suggestions for optimizing learning experience and improving overall performance by making behavioral changes
- Recommendations aimed at enhancing both individual student outcomes and effectiveness of educational system as a whole
- Study highlights importance of analyzing log data in online educational systems to enhance understanding of student learning processes

## Summary

Authors Stephan Lorenzen, Niklas Hjuler, and Stephen Alstrup conducted a study titled "Tracking Behavioral Patterns among Students in an Online Educational System" to gain insights into student learning and improve educational systems. The researchers focused on analyzing log data from Clio Online, the largest provider of digital learning content for primary schools in Denmark. The dataset included information from 14,810 students and 3 million sessions spanning from 2015 to 2017. Using non-negative matrix factorization techniques, the authors uncovered dependencies among various factors such as time of day, subject, activity type, complexity (measured using Bloom's taxonomy), and performance. This comprehensive analysis provided valuable insights into how different aspects of student behavior interact within the educational system. One significant advantage of their method is its ability to track behavioral changes over time for individual students as well as general behavioral changes within the entire educational system. This longitudinal approach allowed for a deeper understanding of how students' behaviors evolve and adapt over time. Based on their findings, the researchers offered suggestions for optimizing the learning experience and improving overall performance by making behavioral changes. The recommendations are aimed at enhancing both individual student outcomes and the effectiveness of the educational system as a whole. Overall, this study highlights the importance of analyzing log data in online educational systems to enhance our understanding of student learning processes. The insights gained from this research have practical implications for improving educational practices and tailoring instruction to meet individual needs effectively.