Geonmin Kim.

I am a master’s student in Artificial Intelligence Convergence at Chonnam National University (DNSLAB, advisor: Prof. Kyungbaek Kim). My primary research identity is AI-based security — intrusion detection, security machine learning, and policy systems for cloud and distributed environments. I also have substantial experience in decentralized identity (DID) — method security, interoperability, and KISA-related work — which complements my interest in trust and verification in distributed systems. Other collaborations (networked AI, applied sensing) appear under Projects and Publications.

Core agenda for doctoral study: making AI-driven security decisions reliable, explainable, and deployable in cloud and distributed systems — with complementary work on decentralized identity trust and interoperability.

1. False-positive-aware federated decision fusion

   Decision-level fusion and thresholding that accounts for asymmetric error costs, so intrusion alerts remain usable under real false-positive constraints.

2. Explainable security policy generation and placement

   SHAP- and transformer-based policy orchestration that turns model evidence into actionable hybrid-cloud security controls and edge–cloud placement.

3. AI-based intrusion detection for cloud and distributed environments

   Learning-based detection and adaptive enforcement across hybrid-cloud and distributed infrastructures, from log-driven models to low-latency policy caching.

4. Decentralized identity (DID) security and interoperability

   Method security analysis, verification consistency, and resolution/interoperability for DID systems — including KISA-related work on document retrieval and trust frameworks.

• Chonnam National University Sep. 2025 – Present

  Master’s student in Artificial Intelligence Convergence

  Advisor: Prof. Kyungbaek Kim · GPA 4.38 / 4.5 · Expected graduation: Aug. 2027 · Gwangju, Republic of Korea

• Chonnam National University Mar. 2020 – Aug. 2025

  Bachelor’s in Software Engineering

  GPA 3.69 / 4.5 · Gwangju, Republic of Korea

  Military service, ROK Air Force — Air Defense & Control Command (Mar. 2021 – Dec. 2022)

English-language documents from Chonnam National University. These redacted public copies are for reference only; personal identifiers have been removed. They are not digitally signed originals or official submission copies.

• Graduate academic transcript

  Artificial Intelligence Convergence · Issued September 22, 2026

  Download redacted transcript (PDF)

• Certificate of coursework completion

  Master’s coursework completed August 26, 2026. This is a certificate of completion, not a degree certificate.

  Download redacted certificate (PDF)

Selected research projects

• AI-Based Source Code Vulnerability Detection

  Industry–Academic Collaboration · Jan. 2025 – Dec. 2025

  CodeBERT- and LoRA-based multi-label vulnerability detection within a national project on EV charging infrastructure security.

• Direct Resolution of DID Documents

  KISA · Jun. 2024 – Dec. 2025

  Foundational and follow-up research on DID method structures, retrieval flows, interoperability, and standardization-oriented documentation.

• AI Diagnosis for Decommissioned Li-Ion Batteries

  Industry–Academic Collaboration · Jun. 2025 – Dec. 2025

  Survival analysis and remaining useful life prediction for decommissioned lithium-ion batteries.

• DID-Based School Access Management

  Industry–Academic Collaboration · Jun. 2024 – Jun. 2025

  Design participation for a DID-based visitor log and identity verification system.

• K-BTF Common Requirements

  KISA · Jul. 2025 – Dec. 2025

  Reviewed interoperability, security, and operational requirements for blockchain and decentralized identity trust frameworks.

• Vision-Based Foreign Material Detection

  Industry–Academic Collaboration · Jul. 2025 – Oct. 2025

  Computer vision inspection model for detecting foreign materials in dried laver production.

49 papers in total, including accepted papers — comprising 3 articles in KCI-indexed journals, 10 international conference papers, and 36 domestic conference papers. First-author publications: 27 (including co-first-authored papers), grouped below by research theme (newest first within each area). Full profile on Google Scholar.

AI security & policy systems

1. International

   FPR-Constrained Explainable Policy Placement for Edge-Cloud Intrusion Detection

   APNOMS 2026.

2. International

   HCAPO: Transformer-Based Adaptive Policy Orchestration for Hybrid-Cloud Security

   ICAIIC 2026.

   PDF Scholar
3. Domestic

   Explainable Policy Generation for Hybrid Intrusion Detection Using SHAP

   ASK 2026 · Best Paper Award.

   PDF
4. International

   Edge-Based Policy Caching for Low Latency Security Enforcement in Hybrid Clouds

   APNOMS 2025 · Best Paper Candidate.

   PDF Scholar
5. Domestic

   CWE Code Vulnerability Detection Using a Lightweight LoRA-Based CodeBERT Model

   ASK 2025 · Best Paper Award (KISA President Award).

   PDF Scholar
6. Domestic

   Multi-Label Vulnerability Detection in Source Code Using CodeBERT

   ASK 2025.

   PDF Scholar
7. Domestic

   Dynamic Security Policy Application and Threat Detection Based on Machine Learning in Hybrid Cloud Environments

   KNOM 2025.

   PDF
8. Domestic

   Design and Validation of a Machine Learning-Based Anomaly Detection Model for Open Banking User Sequences

   IEIE Summer Conference, 2025.

   PDF Scholar
9. Domestic

   Performance Comparison of Machine Learning-Based Web Attack Detection in Cloud Environments

   Korean Institute of Smart Media Fall Conference, 2025.

   PDF
10. Domestic

    Preventive Security Model Construction Based on Time-Series Log Analysis in Hybrid Cloud Environments

    KICS Fall Conference, 2025.

    PDF Scholar
11. Domestic

    Stabilizing Log-Based Security Policy Generation in Hybrid Clouds Through Cross-Selection

    ACK 2025.

    PDF
12. Domestic

    Integrated System for Detecting DNS Spoofing Attacks in Public Network Environments

    2024 Honam Cybersecurity Conference · Best Paper Award.

    PDF
13. Domestic

    A Salting and Steganography-Based Approach for Data Integrity Verification of CoAP in IoT Environments

    ACK 2024 · Best Paper Award (ETRI President Award).

    PDF Scholar

Decentralized identity (DID)

1. Domestic

   Effects of Monitoring Paths and Polling Intervals on Partial Key Revocation Detection in DID web

   Geonmin Kim, Kyungbaek Kim · Korean Institute of Smart Media Fall Conference, 2026 · Accepted.

   PDF
2. Domestic

   Measuring the Effectiveness of Client no-cache Directives in Web-based Decentralized Identifier Resolution

   G. Kim, K. Kim · 2026 Honam Cybersecurity Conference (KIISC Honam Branch).

   PDF
3. Domestic

   Analysis of Inter-Implementation Inconsistencies in DID Identifier Verification and an Error-Classification Preprocessing Module

   G. Kim, M. Choi, S. Jeong, K. Kim · IKEEE Summer Conference 2026.

   PDF
4. Domestic

   A Study on Preventing OpenID Relying Party Phishing Using Blockchain

   KICS Fall Conference, 2024.

   PDF Scholar

Networked AI (traffic / V2X)

1. Domestic

   Minimum Green Time-Based Reinforcement Learning for Multi-Intersection Traffic Signal Control

   Geonmin Kim, Minjun Choi, Kyungbaek Kim · Korea Digital Contents Society Fall Conference, 2026 · Accepted.

   PDF
2. Journal · KCI

   Enhancing Traffic Accident Risk Forecasting via Involution-based Dynamic Spatiotemporal Data Modeling

   Journal of KIISE · Accepted · Forthcoming.

   Geonmin Kim: co-first author (equal contribution).

3. International

   Communication-Aware Adaptive Decentralized Traffic Signal Control via Broadcast Gate

   ICUFN 2026 · pp. 253–259.

   IEEE Xplore DOI
4. Domestic

   Selective Communication Protocol for GNN-Based MARL Traffic Control

   JCCI 2026.

   PDF
5. Domestic

   Learning Value-Based Selective Communication for Traffic Signal Control with Reduced Communication Overhead

   KNOM 2026.

   PDF
6. International

   Comparative Study of 5G-V2X and 6G-V2X for Intelligent Traffic Signal Control in Urban Environment

   SMA 2025.

   PDF
7. Domestic

   MARL-Based Intelligent Traffic Signal Control Model for 6G-V2X

   JCCI 2025.

   PDF

Applied ML & other domains

1. Journal · KCI

   Survival-Analysis Based Prediction of Remaining Useful Life of Lithium-Ion Batteries with an Analysis of Input Feature Effects

   G. Kim et al. · Smart Media Journal, 2026.

   DOI KCI
2. Domestic

   Performance Comparison of Machine Learning Models for Real-Time MBTI Personality Type Classification Based on Natural Language

   Korean Institute of Smart Media Spring Conference, 2025.

   PDF

Selected co-authored

1. Journal · KCI

   Design of a Risk-Budget Reservation and Settlement Layer for Energy Storage Control with Delayed Measurements

   Minjun Choi, Geonmin Kim, Kyungbaek Kim · Smart Media Journal · Accepted · Forthcoming.

2. Domestic

   Horizon-Adaptive Physics-AI Framework for Merge Congestion Length Prediction

   Minjun Choi, Geonmin Kim, Kyungbaek Kim · Korea Digital Contents Society Fall Conference, 2026 · Accepted.

   PDF
3. Domestic

   Adaptive Resolver Fan-out for DID Resolution under Dynamic Resolver Conditions

   Minsu Yoo, Geonmin Kim, Kyungbaek Kim · ACK 2026 · Accepted.

   Best Paper Award (KFITS President's Award) · Awarded November 6, 2026.

   PDF
4. International

   TrizoneMP: Workload-Aware Max-Pressure Evaluation Management for City-Scale Traffic Signal Control

   APNOMS 2026.

5. International

   Analyzing the Benchmark-to-Deployment Threshold Gap in Time-Series Anomaly Detection

   APNOMS 2026.

6. International

   Multi-Agent Reinforcement Learning for Integrated Energy Management in Smart Buildings: A Review and Future Directions

   M. I. Bappi, J. Shin, G. Kim, D. R. Oluwabukola, K. Kim · SMA 2026.

7. International

   Dependency-Informed GraphCodeBERT Framework for Java Method-Level Vulnerability Detection

   U. Shakhakarmi, G. Kim, H. Jang, K. Kim · SMA 2026 · Best Paper Award.

   PDF
8. Domestic

   Does Faithful Data-Flow Access Reduce Lexical Reliance in Java Vulnerability Detection?

   U. Shakhakarmi, H. Jang, G. Kim, K. Kim · 2026 Honam Cybersecurity Conference (KIISC Honam Branch).

   PDF
9. International

   Enhancing Cloud Security with AI Techniques

   S. Yeom, G. Kim, et al. · SMA 2025.

   PDF
10. Domestic

    Lightweight Traffic Congestion Prediction via Hidden-Dimension Reduction for Edge Inference

    M. Choi, G. Kim, S. Jeong, K. Kim · IKEEE Summer Conference 2026.

    PDF
11. Domestic

    Empirical Analysis of Waiting Time and Overhead in Container-Based Hash Search Using Poisson and Gamma Distributions

    M. Yoo, G. Kim, et al. · ASK 2026 · Best Paper Award.

    PDF
12. Domestic

    Physics-Informed Dense Traffic Density Field Reconstruction for Telemetry Resource Optimization

    M. Choi, G. Kim, et al. · KNOM 2026.

    PDF
13. Domestic

    Customer Behavior Prediction Model for Personalized Marketing Based on Machine Learning

    Y. Kim, T. Kim, G. Kim, et al. · KICS Fall Conference, 2025 · Best Paper Award.

    PDF
14. Domestic

    Security Vulnerability Analysis of Decentralized Identifier Methods

    H. Jang, G. Kim, et al. · 2025 Honam Cybersecurity Conference.

    PDF
15. Domestic

    Trend Analysis of Web-Based DID Methods

    D. Lim, G. Kim, et al. · 2024 Honam Cybersecurity Conference.

    PDF
16. arXiv

    Deformable Dynamic Convolution for Accurate yet Efficient Spatio-Temporal Traffic Prediction

    H. Jin, G. Kim, K. Kim · arXiv:2507.11550, 2025.

    arXiv PDF

Complete list in the academic CV (PDF) and on Google Scholar.

• Chonnam National University May 2026 – Present

  Assistant Instructor — SW/AI Mentorship Program

  Science Education Institute for the Gifted. Mentored gifted students in SW/AI research and project presentations.

• Chonnam National University 2025 – 2026

  Teaching Assistant

  Understanding of Blockchain · Fundamentals of Artificial Intelligence (intensive session).

• Invited lectures 2024 – 2025

  Generative AI & Python instruction

  Gwangju Software Meister High School · Jeongup High School.

Academic Awards

1. Excellence Award BK21 FOUR Outstanding Graduate Student Fellowship, 2026
2. Undang Academic Award Korea Information Processing Society, 2024

Best Paper Awards

Includes awards for co-authored papers.

1. Best Paper Award (KFITS President's Award)

   Korea Federation of Information Technology Societies · ACK 2026

   Adaptive Resolver Fan-out for DID Resolution under Dynamic Resolver Conditions

   Co-authored paper: Minsu Yoo, Geonmin Kim, Kyungbaek Kim.

   Awarded November 6, 2026.

   Certificate
2. Best Paper Award SMA 2026 — Dependency-Informed GraphCodeBERT Framework for Java Method-Level Vulnerability Detection
3. Best Paper Award ASK 2026 — Explainable Policy Generation for Hybrid Intrusion Detection Using SHAP
4. Best Paper Award ASK 2026 — Empirical Analysis of Waiting Time and Overhead in Container-Based Hash Search Using Poisson and Gamma Distributions
5. Best Paper Award KICS Fall Conference 2025 — Customer Behavior Prediction Model for Personalized Marketing Based on Machine Learning
6. Best Paper Award SMA 2025 — Enhancing Cloud Security with AI Techniques
7. Best Paper Award (KISA President) ASK 2025 — LoRA-Based CodeBERT vulnerability detection
8. Best Paper Award (ETRI President) ACK 2024 — CoAP integrity via salting & steganography
9. Best Paper Award 2024 Honam Cybersecurity Conference — DNS spoofing detection