본문 바로가기
KERD
논문KCI 등재

검색 증강 생성(RAG) 기술의 최신 연구 동향에 대한 조사

A Survey on the Latest Research Trends in Retrieval-Augmented Generation

배호|이은빈
한국정보처리학회|2024.09|정보처리학회 논문지 (The Transactions of the Korea Information Processing Society)|제13권 제9호|pp. 429-436|DOI 10.3745/TKIPS.2024.13.9.429

2

조회

0

다운로드

1

인용

0

참고문헌

키워드

Hallucination대형 언어 모델검색 증강 생성Retrieval-Augmented GenerationLLM환각 현상
주제분류컴퓨터학

초록

Large Language Model(LLM)의 급격한 발전은 자연어 처리 분야에 혁신을 불러 일으켜 이를 적절하게 활용하는 것이 중요한 주제로 떠오르고있다. 방대한 데이터로 훈련된 LLM은 다양한 주제에 대한 텍스트 생성이 가능하여 콘텐츠 생성, 기계 번역, 챗봇 등 여러 방식으로 적용이 가능하나특정 유형이나 전문적 지식이 부족할 수 있어 일반화하기 어렵다는 단점이 존재한다. 모델 훈련이 완료된 이후의 최신 정보로 즉각 업데이트되기도어려우며, 모델이 실제로 존재하지 않는 정보나 오류에 대해 그럴 듯하게 답변하는 환각 현상(Hallucination) 역시 주요 문제점이다. 이를 극복하기위해 지속적으로 업데이트되는 최신 정보를 포함한 외부 데이터베이스에서 정보를 검색해 응답을 생성하는 Retrieval-Augmented Generation(RAG,검색 증강 생성) 모델을 도입하여 LLM의 환각 현상을 최소화하고 효율성과 정확성을 향상하기 위한 연구가 활발히 이루어지고 있다. 본 논문에서는RAG의 기본 아키텍처를 소개하고, LLM에 RAG를 적용하기 위한 연구 및 최적화의 최신 동향을 분석한다. RAG를 평가하기 위한 다양한 기법들을소개하고, 실제 산업에서 RAG를 활용하기 위해 성능을 최적화하거나 응용한 사례들을 분석한다. 이를 바탕으로 향후 RAG 모델이 발전할 수 있는연구 방향성을 제시하고자 한다.

Abstract

As Large Language Models (LLMs) continue to advance, effectively harnessing their potential has become increasingly important. LLMs,trained on vast datasets, are capable of generating text across a wide range of topics, making them useful in applications such as contentcreation, machine translation, and chatbots. However, they often face challenges in generalization due to gaps in specific or specializedknowledge, and updating these models with the latest information post-training remains a significant hurdle. To address these issues,Retrieval-Augmented Generation (RAG) models have been introduced. These models enhance response generation by retrieving informationfrom continuously updated external databases, thereby reducing the hallucination phenomenon often seen in LLMs while improvingefficiency and accuracy. This paper presents the foundational architecture of RAG, reviews recent research trends aimed at enhancingthe retrieval capabilities of LLMs through RAG, and discusses evaluation techniques. Additionally, it explores performance optimizationand real-world applications of RAG in various industries. Through this analysis, the paper aims to propose future research directionsfor the continued development of RAG models.

참고문헌 (56)

  2. P. Lewis, "Retrieval-augmented generation for knowledge-intensive nlp tasks", Advances in Neural Information Processing Systems, 33, pp.9459~9474 (2020)
  7. S. Robertson, "The probabilistic relevance framework : BM25 and beyond", Foundations and Trends® in Information Retrieval, 3(4), pp.333~389 (2009)
  8. V. Karpukhin, "Dense passage retrieval for open domain question answering", Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing, Online, pp.6769~6781 (2020)
  9. M. Lewis, "Bart: Denoising sequence-to-sequence pre-training for natural language generation, translation, and comprehension", Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, pp.7871~7880 (2020)
  10. G. Izacard, "Leveraging passage retrieval with generative models for open domain question answering", Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics: Main Volume, pp.874~880 (2021)
  13. K. Guu, "Retrieval augmented language model pre-training", Proceedings of the 37th International Conference on Machine Learning, pp.3929~3938 (2020)
  14. D. Chen, "Reading wikipedia to answer open-domain questions", Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics, pp.1870~1879 (2017)
  15. W. Shi, "REPLUG: Retrieval-Augmented Black-Box Language Models", Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pp.8371~8384 (2024)
  17. Z. Jiang, "Active retrieval augmented generation", Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pp.7969~7992 (2023)
  19. Y. Wang, "Self-knowledge guided retrieval augmentation for large language models", Findings of the Association for Computational Linguistics:EMNLP, pp.10303~10315 (2023)
  20. J. Zhang, "ReAugKD : Retrieval-Augmented Knowledge Distillation For Pre-trained Language Models", Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics, pp.1128~1136 (2023)
  22. F. Shi, "Large language models can be easily distracted by irrelevant context", Proceedings of the 40th International Conference on Machine Learning, pp.31210~31227 (2023)
  23. A. Asai, "Evidentiality guided Generation for Knowledge-Intensive NLP Tasks", Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pp.2226~2243 (2022)
  25. X. Ma, "Query Rewriting in Retrieval-Augmented Large Language Models", Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pp.5303~5315 (2023)
  28. J. Saad-Falcon, "ARES: An Automated Evaluation Framework for RetrievalAugmented Generation Systems", Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pp.338~354 (2024)
  30. T. Gao, "Enabling large language models to generate text with citations", Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pp.6465~6488 (2023)
  31. T. Kwiatkowski, "Natural questions : a benchmark for question answering research", Transactions of the Association for Computational Linguistics, pp.452~466 (2019)
  32. A. Mallen, "When Not to Trust Language Models : Investigating Effectiveness of Parametric and Non-Parametric Memories", Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics, pp.9802~9822 (2023)
  33. Z. Yang, "HotpotQA : A dataset for diverse, explainable multi-hop question answering", Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing, pp.2369~2380 (2018)
  34. A. Fan, "ELI5: Long form question answering", Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, pp.3558~3567 (2019)
  35. M. Joshi, "TriviaQA : A Large Scale Distantly Supervised Challenge Dataset for Reading Comprehension", Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics, pp.1601~1611 (2017)
  37. H. Wang, "Large Language Models as Source Planner for Personalized Knowledge-grounded Dialogues", Findings of the Association for Computational Linguistics, pp.9556~9569 (2023)
  38. S. Li, "Document-Level Event Argument Extraction by Conditional Generation", Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pp.894~908 (2021)
  39. S. Ebner, "Multi-Sentence Argument Linking", Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, pp.8057~8077 (2020)
  40. R. Zellers, "HellaSwag: Can a Machine Really Finish Your Sentence?", Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, pp.4791~4800 (2019)
  41. S. Kim, "The CoT Collection : Improving Zero-shot and Few-shot Learning of Language Models via Chainof-Thought Fine-Tuning", Proceedings of the 2023Conference on Empirical Methods in Natural Language Processing, pp.12685~12708 (2023)
  43. Z. Feng, "Retrieval Generation Synergy Augmented Large Language Models", IEEE International Conference on Acoustics, Speech and Signal Processing, pp.11661~11665 (2024)
  44. Z. Shao, "Enhancing Retrieval-Augmented Large Language Models with Iterative Retrieval-Generation Synergy", Findings of the Association for Computational Linguistics, pp.9248~9274 (2023)
  45. Y. Hoshi, "RaLLe: A Framework for Developing and Evaluating Retrieval-Augmented Large Language Models", Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing: System Demonstrations, pp.52~69 (2023)
  46. J. Chen, "Benchmarking Large Language Models in Retrieval-Augmented Generation", Proceedings of the AAAI Conference on Artificial Intelligence, pp.17754~17762 (2024)
  48. A. Lozano, "Clinfo.ai:An Open-Source Retrieval-Augmented Large Language Model System for Answering Medical Questions using Scientific Literature", Pacific Symposium On Biocomputing, pp.8~23 (2024)
  51. W. Chen, "MuRAG : Multimodal Retrieval-Augmented Generator for Open Question Answering over Images and Text", Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, pp.5558~5570 (2022)
  53. Z. Wang, "M-RAG:Reinforcing Large Language Model Performance through Retrieval-Augmented Generation with Multiple Partitions", Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics, pp.1966~1978 (2024)
  54. R. Shrestha, "FairRAG: Fair Human Generation via Fair Retrieval Augmentation", Proceedings of the IEEE, 2024, 11996~12005
  56. E. Lee, "A Survey on Retrieval-Augmented Generation", Proceedings of the Annual Symposium of Korea Information Processing Society Conference (KIPS)2024, 31(1), pp.745~748 (2024)

관련 논문AI추천

질의의 감정과 구체성 특성을 반영한 응답 생성을 위한 검색 증강 생성 기법권준희, 김성림(사)디지털산업정보학회 · (사)디지털산업정보학회 논문지2025 · 제21권 제4호LLM을 활용한 검색 증강 생성(RAG) 모델 구현: 래피드마이너 기반으로양치복, Kim Yang Sok(사)한국스마트미디어학회 · 스마트미디어저널2025 · 제14권 제2호RAG 기반 모델의 성능 향상을 위한 문장화 추출 방법론김남규, 남명국한국컴퓨터정보학회 · 한국컴퓨터정보학회논문지2024 · 제29권 제12호질문 유형과 확신도를 활용한 RAG 호출 최적화 방법김미진, 김채영, 신현학, 이승민, 최재걸한국차세대컴퓨팅학회 · 한국차세대컴퓨팅학회 논문지2025 · 제21권 제5호학사 데이터 기반의 생성형 AI 챗봇 서비스에 관한 연구이준환(사)한국스마트미디어학회 · 스마트미디어저널2024 · 제13권 제12호한국어 법률 QA를 위한 하이브리드 RAG 시스템 최적화민정혜, 서준원한국컴퓨터정보학회 · 한국컴퓨터정보학회논문지2025 · 제30권 제8호

이 간행물의 다른 논문 (9건)

CAN 메시지의 주기성과 시계열 분석을 활용한 비정상 탐지 방법김세린, 성지현, 조학수, 윤범헌pp. 395-403eBPF를 활용한 컨테이너 활동 분석 시스템남재현, 김지수pp. 404-412RISC-V 아키텍처 상에서의 쉐도우 스택 성능 평가 및 고찰강하영, 박성환, 권동현, 한고원pp. 413-420생성형 AI 모델의 한국문화 이해 능력 평가 방법에 관한 연구김승현, 손기준pp. 421-428온라인 평가 시스템에서 코딩 스타일 검사 시스템 구현김영훈, 우균, 천준석pp. 437-443
원문 보기