A Comparative Performance Analysis of Locally Deployed Large Language Models Through a Retrieval-Augmented Generation Educational Assistant Application for Textual Data Extraction

<b>Background:</b> Rapid advancements in large language models (LLMs) have significantly enhanced Retrieval-Augmented Generation (RAG) techniques, leading to more accurate and context-aware information retrieval systems. <b>Methods:</b> This article presents the creation of a...

Full description

Saved in:
Bibliographic Details
Main Authors: Amitabh Mishra, Nagaraju Brahmanapally
Format: Article
Language:English
Published: MDPI AG 2025-06-01
Series:AI
Subjects:
Online Access:https://www.mdpi.com/2673-2688/6/6/119
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849423955756056576
author Amitabh Mishra
Nagaraju Brahmanapally
author_facet Amitabh Mishra
Nagaraju Brahmanapally
author_sort Amitabh Mishra
collection DOAJ
description <b>Background:</b> Rapid advancements in large language models (LLMs) have significantly enhanced Retrieval-Augmented Generation (RAG) techniques, leading to more accurate and context-aware information retrieval systems. <b>Methods:</b> This article presents the creation of a RAG-based chatbot tailored for university course catalogs, aimed at answering queries related to course details and other essential academic information, and investigates its performance by testing it on several locally deployed large language models. By leveraging multiple LLM architectures, we evaluate performance of the models under test in terms of context length, embedding size, computational efficiency, and relevance of responses. <b>Results:</b> The experimental analysis obtained by this research, which builds on recent comparative studies, reveals that while larger models achieve higher relevance scores, they incur greater response times than smaller, more efficient models. <b>Conclusions:</b> The findings underscore the importance of balancing accuracy and efficiency for real-time educational applications. Overall, this work contributes to the field by offering insights into optimal RAG configurations and practical guidelines for deploying AI-powered educational assistants.
format Article
id doaj-art-8925f2ac2ddb43bda54785b9f9a5cf87
institution Kabale University
issn 2673-2688
language English
publishDate 2025-06-01
publisher MDPI AG
record_format Article
series AI
spelling doaj-art-8925f2ac2ddb43bda54785b9f9a5cf872025-08-20T03:30:24ZengMDPI AGAI2673-26882025-06-016611910.3390/ai6060119A Comparative Performance Analysis of Locally Deployed Large Language Models Through a Retrieval-Augmented Generation Educational Assistant Application for Textual Data ExtractionAmitabh Mishra0Nagaraju Brahmanapally1Department of Computer Science, University of West Florida, 11000 University Parkway, Pensacola, FL 32514, USADepartment of Computer Science, University of West Florida, 11000 University Parkway, Pensacola, FL 32514, USA<b>Background:</b> Rapid advancements in large language models (LLMs) have significantly enhanced Retrieval-Augmented Generation (RAG) techniques, leading to more accurate and context-aware information retrieval systems. <b>Methods:</b> This article presents the creation of a RAG-based chatbot tailored for university course catalogs, aimed at answering queries related to course details and other essential academic information, and investigates its performance by testing it on several locally deployed large language models. By leveraging multiple LLM architectures, we evaluate performance of the models under test in terms of context length, embedding size, computational efficiency, and relevance of responses. <b>Results:</b> The experimental analysis obtained by this research, which builds on recent comparative studies, reveals that while larger models achieve higher relevance scores, they incur greater response times than smaller, more efficient models. <b>Conclusions:</b> The findings underscore the importance of balancing accuracy and efficiency for real-time educational applications. Overall, this work contributes to the field by offering insights into optimal RAG configurations and practical guidelines for deploying AI-powered educational assistants.https://www.mdpi.com/2673-2688/6/6/119large language modelsGenerative Artificial IntelligenceRetrieval-Augmented Generationnatural language processing
spellingShingle Amitabh Mishra
Nagaraju Brahmanapally
A Comparative Performance Analysis of Locally Deployed Large Language Models Through a Retrieval-Augmented Generation Educational Assistant Application for Textual Data Extraction
AI
large language models
Generative Artificial Intelligence
Retrieval-Augmented Generation
natural language processing
title A Comparative Performance Analysis of Locally Deployed Large Language Models Through a Retrieval-Augmented Generation Educational Assistant Application for Textual Data Extraction
title_full A Comparative Performance Analysis of Locally Deployed Large Language Models Through a Retrieval-Augmented Generation Educational Assistant Application for Textual Data Extraction
title_fullStr A Comparative Performance Analysis of Locally Deployed Large Language Models Through a Retrieval-Augmented Generation Educational Assistant Application for Textual Data Extraction
title_full_unstemmed A Comparative Performance Analysis of Locally Deployed Large Language Models Through a Retrieval-Augmented Generation Educational Assistant Application for Textual Data Extraction
title_short A Comparative Performance Analysis of Locally Deployed Large Language Models Through a Retrieval-Augmented Generation Educational Assistant Application for Textual Data Extraction
title_sort comparative performance analysis of locally deployed large language models through a retrieval augmented generation educational assistant application for textual data extraction
topic large language models
Generative Artificial Intelligence
Retrieval-Augmented Generation
natural language processing
url https://www.mdpi.com/2673-2688/6/6/119
work_keys_str_mv AT amitabhmishra acomparativeperformanceanalysisoflocallydeployedlargelanguagemodelsthrougharetrievalaugmentedgenerationeducationalassistantapplicationfortextualdataextraction
AT nagarajubrahmanapally acomparativeperformanceanalysisoflocallydeployedlargelanguagemodelsthrougharetrievalaugmentedgenerationeducationalassistantapplicationfortextualdataextraction
AT amitabhmishra comparativeperformanceanalysisoflocallydeployedlargelanguagemodelsthrougharetrievalaugmentedgenerationeducationalassistantapplicationfortextualdataextraction
AT nagarajubrahmanapally comparativeperformanceanalysisoflocallydeployedlargelanguagemodelsthrougharetrievalaugmentedgenerationeducationalassistantapplicationfortextualdataextraction