A Conversational AI Agent System Based on Large Language Models and Retrieval-Augmented Generation to Support Investigation
Journal of Data Forensics Research · 31.12.2024 · 1 (1) · 47-62
DOI: 10.12972/jdfr.2024.1.1.4 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Data Forensics Research, online erschienen am 31.12.2024. Verfasst von Jeehyeong Kim und 6 weiteren Autor:innen.
Abstract (Originalsprache)
Criminal investigation is a field that requires a high level of expertise and a thorough understanding of various legal frameworks. Thus, there is a growing need for an artificial intelligence (AI) agent service that can effectively provide knowledge-based Q&A support within investigative contexts. This study proposes a reference architecture and benchmark for evaluating the performance of an AI chatbot system utilizing Large Language Models (LLM) and Retrieval-Augmented Generation (RAG). Through this, we offer practical guidelines for developing LLM-based AI systems specifically tailored to investigative work. Additionally, by applying and testing recent models such as GPT-4o and EXAONE 3.0, we demonstrate the feasibility of developing practical, field-ready AI systems for this domain.
Zitieren
APA
Kim, J., Kim, M., Kye, H., Choi, W., Lee, S., Ko, J., & Lee, S. (2024). A Conversational AI Agent System Based on Large Language Models and Retrieval-Augmented Generation to Support Investigation. Journal of Data Forensics Research, 1(1), 47-62. https://doi.org/10.12972/jdfr.2024.1.1.4
Vancouver
Kim J, Kim M, Kye H, Choi W, Lee S, Ko J, et al. A Conversational AI Agent System Based on Large Language Models and Retrieval-Augmented Generation to Support Investigation. Journal of Data Forensics Research. 2024;1(1):47-62. doi:10.12972/jdfr.2024.1.1.4
BibTeX
@article{Kim2024,
author = {Kim, Jeehyeong and Kim, Minsu and Kye, Hyoseon and Choi, Won-Gi and Lee, Seungwoo and Ko, Jaejin and Lee, Sang-Shin},
title = {A Conversational AI Agent System Based on Large Language Models and Retrieval-Augmented Generation to Support Investigation},
journal = {Journal of Data Forensics Research},
year = {2024},
volume = {1},
number = {1},
pages = {47--62},
doi = {10.12972/jdfr.2024.1.1.4}
}
EUVIMED