Data mining student activity patterns in an interactive activity-based STEM learning environment
No Thumbnail Available
Authors
Meeting name
Sponsors
Date
Journal Title
Format
Thesis
Subject
Abstract
Jupyter Notebook is gaining in popularity for STEM instruction and activity-based learning. This platform for sharing interactive documents via a web interface allows instructors to combine a variety of media together with interactive and editable code, providing rich opportunities for an active learning pedagogy. Other online learning environments, such as Canvas and Moodle, provide or integrate learning analytics for the use of administrators, educators, and students to improve learning outcomes; however, these platforms lack the rich learning environment of Jupyter Notebook. Also, with increasing interest in online learning, research communities have arisen for Learning Analytics and Educational Data Mining. Unfortunately, these research communities have not yet begun to address the Jupyter Notebook learning environment. The University of Missouri College of Engineering offers a Program of Study in Data Science (PSDS) under contract with the National Geospatial Intelligence Agency (NGA.) This program is delivered online, making heavy use of Jupyter notebooks served by JupyterHub for active engagement with course content. The PSDS infrastructure uses the Graylog log management program to collect Jupyter logs, which are stored in an integrated Elasticsearch document store for a period of months. The PSDS program provides an excellent case study for a proof-of-concept in applying learning analytics to the Jupyter learning environment. This thesis consists of two major parts. (1) Mining the Graylog system to extract useful log messages, transformation of those messages into student-activities features, and loading the data into a PostgreSQL database for long-term storage. (2) Developing a variety of visualizations of student activity for administrators, instructors and students. The pedological structure of PSDS courses allows unique insights into student engagement with the course material. Finally, recommendations are made for the development of a more comprehensive logging system and additional analyses that could be performed.
Table of Contents
DOI
PubMed ID
Degree
M.S.
Thesis Department
Rights
OpenAccess.
License
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 License. Copyright held by author.