Scholarlink

The scope of this paper is to analyze the web analytics of a Next Generation Digital Learning Environment (NGDLE), which was part of a European initiative, specifically focusing on how platform engagement was affected by the COVID-19 pandemic waves and the mandatory e-schooling in Europe. Emphasis was given to the Bounce Rate (%) variable, a metric to estimate the platform’s effectiveness in retaining its users. This was achieved by applying a classification problem, comparing different machine learning models, and identifying the most contributing features. The study revealed patterns in the data, providing an overview of user engagement in the e-service while unfolding the complex relationships among the features and the need for robust analysis. Finally, it highlighted features contributing to the target variable Bounce Rate (%) that could be further explored to enhance the VLE.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Analysing User Engagement in e-Learning Platforms: A Machine Learning Approach to Web Analytics

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Gouvatsou Christina,
  • Dimitris Pantazatos,
  • Mary Grammatikou,
  • Vasilis Maglaris,
  • Petros Stefaneas

摘要

The scope of this paper is to analyze the web analytics of a Next Generation Digital Learning Environment (NGDLE), which was part of a European initiative, specifically focusing on how platform engagement was affected by the COVID-19 pandemic waves and the mandatory e-schooling in Europe. Emphasis was given to the Bounce Rate (%) variable, a metric to estimate the platform’s effectiveness in retaining its users. This was achieved by applying a classification problem, comparing different machine learning models, and identifying the most contributing features. The study revealed patterns in the data, providing an overview of user engagement in the e-service while unfolding the complex relationships among the features and the need for robust analysis. Finally, it highlighted features contributing to the target variable Bounce Rate (%) that could be further explored to enhance the VLE.