An LLM-Assisted Decision Support System for Automated CLO-PLO Mapping and Academic Evaluation Reporting
Year: 2026
Type: conference-paper
Abstract: This study presents a web-based system that systematically establishes the relationship between Course Learning Outcomes (CLOs) and Program Learning Outcomes (PLOs) and reports student competency levels, with the aim of improving quality assurance processes in higher education institutions. Targeting compliance with national and international standards such as the Bologna Process and the National Qualifications Framework for Higher Education in Turkey (NQF-HETR), the system addresses the time-consuming and inconsistent nature of traditional manual and subjective assessment methods. One of its key innovations is the automatic mapping of exam questions, CLOs, and PLOs, which directly uses unprocessed student performance data and transforms the results into actionable analytical reports. Crucially, the system integrates an evaluation layer supported by Large Language Models (LLMs), including gpt-4o and local models, to interpret outputs and generate structured recommendations. This LLM integration converts static data into dynamic pedagogical insights by providing targeted guidance to academic staff for curriculum adjustments and improvement actions. Built on an ASP.NET Core MVC and Microsoft SQL Server architecture, the system was piloted at Kocaeli University, Türkiye, where its technical feasibility and scalability were validated. Consequently, the developed framework enhances institutional reporting transparency and accountability, shifting learning-outcome assessment from a compliance task to a dynamic, LLM-assisted decision-support mechanism for continuous improvement.
Institution: Kocaeli Üniversitesi
Language: English
Subfield: Education
Field: Social Sciences
Domain: Social Sciences
SDG: Quality education
Open Access status: hybrid
OpenAlex ID: W7166554308