Evaluating the Maturity and Stress Resilience of Learning Management Systems
Jamil, Asma (2025)
2025
Tietojenkäsittelyopin maisteriohjelma - Master's Programme in Computer Science
Informaatioteknologian ja viestinnän tiedekunta - Faculty of Information Technology and Communication Sciences
This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Hyväksymispäivämäärä
2025-06-30
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:tuni-202506307461
https://urn.fi/URN:NBN:fi:tuni-202506307461
Tiivistelmä
The rapid expansion of e-learning in higher education has intensified the reliance on Learning Management Systems (LMS) to deliver, manage, and assess instruction. However, the concurrent use of multiple LMS platforms within institutions has led to fragmentation in course delivery, negatively impacting student learning continuity and increasing the workload for educators who must adapt and manage content across different systems. This thesis investigates the implications of this multi-platform environment and aims to identify a unified, scalable LMS solution that can streamline educational processes and reduce redundancy in instructional design. In addition to platform fragmentation, educators face increasing pressure due to large student enrollments and the time-intensive nature of grading, particularly in programming and data-centric disciplines. As a response, this study explores the potential of AI-based auto-grading tools to alleviate grading burdens and enhance feedback efficiency.
The research includes a qualitative thematic analysis of stakeholder interviews conducted at Tampere University, focusing on four grading platforms—Weto, Plussa, Moodle, and CodeGrade. Findings reveal that while Weto and Plussa offer advanced functionality for programming education, they present challenges related to scalability, usability, and maintenance. In contrast, Moodle, supported institutionally, provides stable infrastructure but limited automation capabilities. CodeGrade emerges as a user-friendly, integrative solution for scalable AI-enhanced assessment. Based on these findings, the study suggests further exploration of the Moodle + CodeGrade combination as a unified LMS solution for Tampere University, given its potential to consolidate systems, reduce instructional burden, and support scalable, modern assessment workflows.
The study also involves a technical evaluation of AI auto-grading tools, implemented using OpenAI’s GPT-4.1 model within Plussa and CodeGrade. Results indicate strong alignment between AI and human grading in well-defined tasks, but limitations persist in complex scenarios requiring contextual interpretation and nuanced feedback. Consistent with insights from the Digivisio 2030 podcast series, the findings advocate for using AI primarily in formative assessment, where automated feedback can augment instructional efficiency without compromising fairness or pedagogical integrity.
Ultimately, this thesis concludes that AI has high potential to improve assessment workflows in digital education. However, its role should be limited to supporting formative learning processes, with human educators maintaining responsibility for summative evaluations to en-sure ethical, transparent, and educationally sound practices.
The research includes a qualitative thematic analysis of stakeholder interviews conducted at Tampere University, focusing on four grading platforms—Weto, Plussa, Moodle, and CodeGrade. Findings reveal that while Weto and Plussa offer advanced functionality for programming education, they present challenges related to scalability, usability, and maintenance. In contrast, Moodle, supported institutionally, provides stable infrastructure but limited automation capabilities. CodeGrade emerges as a user-friendly, integrative solution for scalable AI-enhanced assessment. Based on these findings, the study suggests further exploration of the Moodle + CodeGrade combination as a unified LMS solution for Tampere University, given its potential to consolidate systems, reduce instructional burden, and support scalable, modern assessment workflows.
The study also involves a technical evaluation of AI auto-grading tools, implemented using OpenAI’s GPT-4.1 model within Plussa and CodeGrade. Results indicate strong alignment between AI and human grading in well-defined tasks, but limitations persist in complex scenarios requiring contextual interpretation and nuanced feedback. Consistent with insights from the Digivisio 2030 podcast series, the findings advocate for using AI primarily in formative assessment, where automated feedback can augment instructional efficiency without compromising fairness or pedagogical integrity.
Ultimately, this thesis concludes that AI has high potential to improve assessment workflows in digital education. However, its role should be limited to supporting formative learning processes, with human educators maintaining responsibility for summative evaluations to en-sure ethical, transparent, and educationally sound practices.