SENIOR RESEARCHER KISTI(Korea Institute of Science and Technology Information) | 2020.12 – present
Senior researcher & developer for LLM • Develop science and technology information specialized LLM, KONI • Develop R&D Specialized LLM, KoRnDAlpaca, and apply the model to NTIS • Develop AI algorithm for identifying company names • Develop Classification model and chatbot for NTIS
• Develop patent-based similar company recommendation model
• Develop R&D report auto-generation model
• Develop hadoop-based big data platform and data integration automation module
RESEARCHER KEPRI (KEPCO Research Institute, KEPRI) | 2018.06 – 2020.12
Industrial data scientist & data engineer • Develop cloud platform based on Openstack and Kubernetes
• Develop AI APIs such as OCR, TTS, and custom vision
• Develop automl module specialized for electric industry
• Develop deep learning algorithm for equipment verification
RESEARCHER KAIST | 2018.02 – 2018.06
Research on distributed deep learning and federated learning
•AI for Science( January 2026 - present)
• Research on Superintelligence Technology for Science and Technology Information (January 2022 – December 2025)
• Development of Core Technologies for Intelligent R&D Information Services Based on Big Data (January 2021 – December 2025)
• National Science and Technology Knowledge Information Service Project (January 2021 – December 2025)
• KEPCO Software Development Cloud Platform (July 2018 – December 2022)
• Development of KEPCO Legal Expert System (December 2018 – November 2019)
• Development of Automatic Meter Reading Algorithm Based on YOLO and CNN (August 2018 – December 2019)
• Data Analysis of IDPP (Intelligent Digital Power Plant) PI (February 2019)
• Collaborative Giga-class Smart Cloudlet Core Technology Development (March 2016 – February 2018)
• Study on Presale Price and Presale Rate Prediction Model Based on Real Estate Data (May 2016 – April 2017)
• KISTI Excellent Early-Career Researcher Award (2022)
• NST Excellent Early-Career Researcher Award (2025)
• Advisory Committee Member, Big Data Statistics Development Council, National Data Agency (June 2025–present)
• AI Research Advisory Committee Member, National Statistical Research Institute, National Data Agency (Dec 2025–present)
• Jang, Gwangseon, Hongseok Choi, Chanuk Lim, Kyong-Ha Lee, and Mun Yong Yi. "Leveraging Pretrained Knowledge at Inference Time: LoRA-Gated Contrastive Decoding for Multilingual Factual Language Generation in Adapted LLMs." Proceedings of the Fourteenth International Conference on Learning Representations (ICLR) (2026).
•Jang, G., Kim, M.W., Nam, Yj. et al. EnergyRoute: energy-based uncertainty routing for selective retrieval and large language model assistance in the hierarchical classification of biotechnology R&D projects. Sci Rep (2026) (SCIE Q1)
• Jang, Gwangseon, Hyeon Ji Jeong, and Mun Yong Yi. "MICD: More intra-class diversity in few-shot text classification with many classes." Knowledge-Based Systems (2024): 112851. (SCIE, Q1)
• Jang, Gwangseon & Kim, Yunjeong & Hwang, Myeong-Ha. (2023). SMART-ID: Semantic Matching based on Applying BERT for R&D Institution Identification. The Transactions of The Korean Institute of Electrical Engineers. 72. 1097-1106. 10.5370/KIEE.2023.72.9.1097. (SCOPUS)
• Jeong, H. J., Jang, G., Shin, D., & Kim, T. H. (2022). Automatic Linkage Model of Classification Systems Based on a Pretraining Language Model for Interconnecting Science and Technology with Job Information. Journal of Information Science Theory and Practice, 0(S), 39–45. https://doi.org/10.1633/JISTAP.2022.10.S.4 (KCI)
• Jang, Gwangseon & Hwang, Myeong-Ha. (2021). Automated Machine Learning Pipeline System Based on Beam Search for Electric Power Industry. Transactions of the Korean Institute of Electrical Engineers. 70. 1914-1923. 10.5370/KIEE.2021.70.12.1914. (SCOPUS)
• Jang, Gwangseon & Lee, Jin-woo & Lee, Jae-Gil & Liu, Yunxin. (2020). Distributed fine-tuning of CNNs for image retrieval on multiple mobile devices. Pervasive and Mobile Computing. 64. 101134. 10.1016/j.pmcj.2020.101134. (SCIE, Q2)
• Lee, Jin-woo & Jang, Gwangseon & Jung, Hohyun & Lee, Jae-Gil & Lee, Uichin. (2019). Maximizing MapReduce job speed and reliability in the mobile cloud by optimizing task allocation. Pervasive and Mobile Computing. 60. 101082. 10.1016/j.pmcj.2019.101082. (SCIE, Q2)
(+82) 042 869 1735 [email protected]
[www.linkedin.com/in/광선-장-a5a16bb6/](https://eight-hydrant-1af.notion.site/www-linkedin-com-in-a5a16bb6-13f0303e2652814e833edbb175c029a7)
[github.com/rayjang](https://eight-hydrant-1af.notion.site/github-com-rayjang-13f0303e265281588ed5fc91d3823424)
KAIST Ph.D in Graduate School of Data Science 2022 – 2026
KAIST MS in Graduate School of Knowledge Service Engineering 2016 – 2018
HANGDONG GLOBAL UNIVERSITY BS in Computer Science & Economics 2009 – 2016
• Deep learning & Machine learning • Data engineering
• LLM
• NLP
• Computer vision • Federated learning
• Pytorch
• PySpark
• Openstack & kubernetes
• Korean - native • English