An Exploratory Study on a RAG-based Question-Answering System for University Regulations
Abstract
This study proposes a Retrieval-Augmented Generation (RAG) based QA system to overcome the limitations of conventional keyword-based search in university regulation retrieval. The system integrates a hybrid search strategy with re-ranking to enhance retrieval precision and employs citation-based prompting to ensure response reliability. Quantitative evaluation using the RAGAS framework on 30 representative queries yielded an average Answer Relevancy of 0.87 and Faithfulness of 0.91. These results demonstrate that the proposed system generates reliable, regulation-grounded responses with enhanced accessibility.
Fields of Study & Subjects
Topic Modeling
Information Retrieval and Search Behavior
Natural Language Processing Techniques
Publication details
- Journal / source
- Korean Institute of Smart Media
- Volume
- 15
- Issue
- 2
- Pages
- 151–164
- Publication date
- 27 February 2026
- Language
- EN