@article{ART003283783},
author={Ji Soo Ryu and Jung, Soon Ki},
title={Empirical Study on 70B-Level AI Interview Evaluation Performance of Small LLaMA 3.1 8B Using Generative AI and LLM Techniques},
journal={Journal of Internet of Things and Convergence},
issn={2466-0078},
year={2025},
volume={11},
number={6},
pages={28}
TY - JOUR
AU - Ji Soo Ryu
AU - Jung, Soon Ki
TI - Empirical Study on 70B-Level AI Interview Evaluation Performance of Small LLaMA 3.1 8B Using Generative AI and LLM Techniques
JO - Journal of Internet of Things and Convergence
PY - 2025
VL - 11
IS - 6
PB - The Korea Internet of Things Society
SP - 28
EP -
SN - 2466-0078
AB - This study proposes an integrated AI interviewer system framework for on-premise environments. The system combines an STT–LLM pipeline, a Rasa-based dialogue manager, and a KoBERT–MeCab preprocessing module to enable real-time interview dialogue processing, automated evaluation, and feedback refinement. Applicants’ spoken responses are converted via STT, after which the LLM functions as an interviewer and evaluates responses based on professional relevance, logical consistency, coherence, and feedback usefulness. The application of Retrieval-Augmented Generation (RAG) and Reflection Tuning improves response quality and evaluation consistency. Experimental results show that the LLaMA 3.1 8B model with the proposed architecture demonstrates competitive performance compared to the 70B model under identical conditions, supporting its applicability from a non-inferiority perspective. These findings suggest that small-scale language models can achieve performance comparable to larger models through appropriate architectural design.
KW - AI Interviewer;On-Premise Artificial Intelligence;Speech Recognition-Based Interview;;Retrieval-Augmented Generation (RAG);Automated Interview Response Evaluation
DO -
UR -
ER -
Ji Soo Ryu and Jung, Soon Ki. (2025). Empirical Study on 70B-Level AI Interview Evaluation Performance of Small LLaMA 3.1 8B Using Generative AI and LLM Techniques. Journal of Internet of Things and Convergence, 11(6), 28.
Ji Soo Ryu and Jung, Soon Ki. 2025, "Empirical Study on 70B-Level AI Interview Evaluation Performance of Small LLaMA 3.1 8B Using Generative AI and LLM Techniques", Journal of Internet of Things and Convergence, vol.11, no.6 28.
Ji Soo Ryu, Jung, Soon Ki "Empirical Study on 70B-Level AI Interview Evaluation Performance of Small LLaMA 3.1 8B Using Generative AI and LLM Techniques" Journal of Internet of Things and Convergence 11.6 28 (2025) : 28.
Ji Soo Ryu, Jung, Soon Ki. Empirical Study on 70B-Level AI Interview Evaluation Performance of Small LLaMA 3.1 8B Using Generative AI and LLM Techniques. 2025; 11(6), 28.
Ji Soo Ryu and Jung, Soon Ki. "Empirical Study on 70B-Level AI Interview Evaluation Performance of Small LLaMA 3.1 8B Using Generative AI and LLM Techniques" Journal of Internet of Things and Convergence 11, no.6 (2025) : 28.
Ji Soo Ryu; Jung, Soon Ki. Empirical Study on 70B-Level AI Interview Evaluation Performance of Small LLaMA 3.1 8B Using Generative AI and LLM Techniques. Journal of Internet of Things and Convergence, 11(6), 28.
Ji Soo Ryu; Jung, Soon Ki. Empirical Study on 70B-Level AI Interview Evaluation Performance of Small LLaMA 3.1 8B Using Generative AI and LLM Techniques. Journal of Internet of Things and Convergence. 2025; 11(6) 28.
Ji Soo Ryu, Jung, Soon Ki. Empirical Study on 70B-Level AI Interview Evaluation Performance of Small LLaMA 3.1 8B Using Generative AI and LLM Techniques. 2025; 11(6), 28.
Ji Soo Ryu and Jung, Soon Ki. "Empirical Study on 70B-Level AI Interview Evaluation Performance of Small LLaMA 3.1 8B Using Generative AI and LLM Techniques" Journal of Internet of Things and Convergence 11, no.6 (2025) : 28.
