Detecting Learning Behavior in Programming Assignments by Analyzing Versioned Repositories
Loading...
Access rights
openAccess
CC BY-NC-ND
CC BY-NC-ND
publishedVersion
URL
Journal Title
Journal ISSN
Volume Title
A1 Alkuperäisartikkeli tieteellisessä aikakauslehdessä
This publication is imported from Aalto University research portal.
View publication in the Research portal (opens in new window)
View/Open full text file from the Research portal (opens in new window)
Other link related to publication (opens in new window)
View publication in the Research portal (opens in new window)
View/Open full text file from the Research portal (opens in new window)
Other link related to publication (opens in new window)
Date
Department
Major/Subject
Mcode
Degree programme
Language
en
Pages
17
Series
IEEE Access, Volume 12, pp. 188828-188844
Abstract
Computing education plays a significant role in shaping the calibre of future computing professionals; hence, improving its quality is a valuable endeavour. A promising approach to enhance computing education is leveraging student data from version control systems (VCS). While previous studies have utilised VCS data to predict academic performance, there remains a gap in harnessing this data for learning analytics to understand student learning behaviours in real time. In this research, we introduce the Polivr ecosystem, a comprehensive platform designed to address this gap by utilising VCS data for learning analytics in computing education. The Polivr ecosystem comprises three key modules: Polivr Anonymiser, which ensures data privacy by anonymising student identities; Polivr Core, which mines learning metrics from Git repositories; and Polivr Web Viewer, which transforms the raw metrics into insightful visualisations for educators. We evaluated Polivr using anonymised repositories collected from undergraduate computing courses. The resulting visualisations revealed trends and patterns in student learning behaviours, such as coding habits and progression over time. These insights provide valuable information for educators to enhance teaching strategies and potentially identify at-risk students. This research demonstrates the potential of version control systems as a rich source of learning analytics, contributing to improving computing education by enabling data-driven decision-making in instructional design and student support.Description
Other note
Citation
Chen, J, Lau, S, Leinonen, J, Terragni, V & Giacaman, N 2024, 'Detecting Learning Behavior in Programming Assignments by Analyzing Versioned Repositories', IEEE Access, vol. 12, pp. 188828-188844. https://doi.org/10.1109/ACCESS.2024.3514843