Investigating Flavors of RAG for Applications in College Chatbots

Christian Sarmiento, Eitel Lauría

2025

Abstract

Retrieval Augmented Generation (RAG) has become a growing area of interest in machine learning (ML) and large language models (LLM) for its ability to improve reasoning by grounding responses in relevant contexts. This study analyzes two RAG architectures, RAG’s original design and Corrective RAG. Through a detailed examination of these architectures, their components, and performance, this work underscores the need for robust metrics when assessing RAG architectures and highlights the importance of good quality context documents in building systems that can mitigate LLM limitations, providing valuable insight for academic institutions to design efficient and accurate question-answering systems tailored to institutional needs.

Paper Citation

in Harvard Style

Sarmiento C. and Lauría E. (2025). Investigating Flavors of RAG for Applications in College Chatbots. In Proceedings of the 17th International Conference on Computer Supported Education - Volume 2: CSEDU; ISBN 978-989-758-746-7, SciTePress, pages 421-428. DOI: 10.5220/0013468200003932

in Bibtex Style

@conference{csedu25,
author={Christian Sarmiento and Eitel Lauría},
title={Investigating Flavors of RAG for Applications in College Chatbots},
booktitle={Proceedings of the 17th International Conference on Computer Supported Education - Volume 2: CSEDU},
year={2025},
pages={421-428},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013468200003932},
isbn={978-989-758-746-7},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 17th International Conference on Computer Supported Education - Volume 2: CSEDU
TI - Investigating Flavors of RAG for Applications in College Chatbots
SN - 978-989-758-746-7
AU - Sarmiento C.
AU - Lauría E.
PY - 2025
SP - 421
EP - 428
DO - 10.5220/0013468200003932
PB - SciTePress