Investigating the Performance of Retrieval-Augmented Generation and Domain-Specific Fine-Tuning for the Development of AI-Driven Knowledge-Based Systems

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Machine Learning and Knowledge Extraction vol. 7, no. 1 (2025), p. 15
1. Verfasser: Lakatos, Róbert
Weitere Verfasser: Pollner, Péter, Hajdu, András, Joó, Tamás
Veröffentlicht:
MDPI AG
Schlagworte:
Online-Zugang:Citation/Abstract
Full Text + Graphics
Full Text - PDF
Tags: Tag hinzufügen
Keine Tags, Fügen Sie das erste Tag hinzu!

MARC

LEADER 00000nab a2200000uu 4500
001 3181641098
003 UK-CbPIL
022 |a 2504-4990 
024 7 |a 10.3390/make7010015  |2 doi 
035 |a 3181641098 
045 2 |b d20250101  |b d20250331 
100 1 |a Lakatos, Róbert  |u Department of Data Science and Visualization, Faculty of Informatics, University of Debrecen, 4032 Debrecen, Hungary; <email>hajdu.andras@inf.unideb.hu</email>; Doctoral School of Informatics, University of Debrecen, 4032 Debrecen, Hungary; Neumann Technology Platform, Neumann Nonprofit Ltd., 1074 Budapest, Hungary 
245 1 |a Investigating the Performance of Retrieval-Augmented Generation and Domain-Specific Fine-Tuning for the Development of AI-Driven Knowledge-Based Systems 
260 |b MDPI AG  |c 2025 
513 |a Journal Article 
520 3 |a Generative large language models (LLMs) have revolutionized the development of knowledge-based systems, enabling new possibilities in applications like ChatGPT, Bing, and Gemini. Two key strategies for domain adaptation in these systems are Domain-Specific Fine-Tuning (DFT) and Retrieval-Augmented Generation (RAG). In this study, we evaluate the performance of RAG and DFT on several LLM architectures, including GPT-J-6B, OPT-6.7B, LLaMA, and LLaMA-2. We use the ROUGE, BLEU, and METEOR scores to evaluate the performance of the models. We also measure the performance of the models with our own designed cosine similarity-based Coverage Score (CS). Our results, based on experiments across multiple datasets, show that RAG-based systems consistently outperform those fine-tuned with DFT. Specifically, RAG models outperform DFT by an average of 17% in ROUGE, 13% in BLEU, and 36% in CS. At the same time, DFT achieves only a modest advantage in METEOR, suggesting slightly better creative capabilities. We also highlight the challenges of integrating RAG with DFT, as such integration can lead to performance degradation. Furthermore, we propose a simplified RAG-based architecture that maximizes efficiency and reduces hallucination, underscoring the advantages of RAG in building reliable, domain-adapted knowledge systems. 
651 4 |a United States--US 
653 |a Language 
653 |a Work stations 
653 |a Accuracy 
653 |a Datasets 
653 |a Performance evaluation 
653 |a Large language models 
653 |a Knowledge 
653 |a Benchmarks 
653 |a Retrieval 
653 |a Machine translation 
653 |a Performance degradation 
653 |a Coronaviruses 
653 |a Chatbots 
653 |a Meteors 
700 1 |a Pollner, Péter  |u Data-Driven Health Division of National Laboratory for Health Security, Health Services Management Training Centre, Semmelweis University, 1085 Budapest, Hungary 
700 1 |a Hajdu, András  |u Department of Data Science and Visualization, Faculty of Informatics, University of Debrecen, 4032 Debrecen, Hungary; <email>hajdu.andras@inf.unideb.hu</email> 
700 1 |a Joó, Tamás  |u Neumann Technology Platform, Neumann Nonprofit Ltd., 1074 Budapest, Hungary; Data-Driven Health Division of National Laboratory for Health Security, Health Services Management Training Centre, Semmelweis University, 1085 Budapest, Hungary 
773 0 |t Machine Learning and Knowledge Extraction  |g vol. 7, no. 1 (2025), p. 15 
786 0 |d ProQuest  |t Advanced Technologies & Aerospace Database 
856 4 1 |3 Citation/Abstract  |u https://www.proquest.com/docview/3181641098/abstract/embedded/H09TXR3UUZB2ISDL?source=fedsrch 
856 4 0 |3 Full Text + Graphics  |u https://www.proquest.com/docview/3181641098/fulltextwithgraphics/embedded/H09TXR3UUZB2ISDL?source=fedsrch 
856 4 0 |3 Full Text - PDF  |u https://www.proquest.com/docview/3181641098/fulltextPDF/embedded/H09TXR3UUZB2ISDL?source=fedsrch