[The Journal of the Korea Society for Naval Science & Technology: Real-time Threat Analysis and Decision Support System for the Korean Peninsula Based on Multi-AI Model Fusion]
- Paul
- Oct 19
- 1 min read
My research paper "Real-time Threat Analysis and Decision Support System for the Korean Peninsula Based on Multi-AI Model Fusion" has been published in the Journal of the Korea Society for Naval Science & Technology on September 30th.
As a Navy reservist, it is deeply meaningful to contribute, even in a small way, to the ROK Navy and national defense. I have invested nearly a year in planning and developing TARGIS BUSAN, continuing to advance it even on holidays and weekends.
The Korean Peninsula faces complex threats including North Korea's nuclear and missile programs and China's military buildup, requiring rapid threat assessment within KADIZ. However, existing single-domain surveillance systems have shown limitations.
To address this, I integrated cutting-edge language models like OpenAI GPT-5 and Anthropic Claude-4 with proven algorithms such as Kalman filters and Bayesian classification. The system integrates GPT-5 and Claude-4 securely, implementing 4-level risk analysis and 6 scenario-based response protocols.
I operated the system using real military base data from 6 countries (South Korea, North Korea, China, Japan, Russia, and the United States) - covering over 500 Army, Air Force, Navy, and Missile units - combined with civilian aircraft data.
TARGIS BUSAN will continue to evolve. Those interested in joint research or collaboration are welcome to reach out.
Korea Society for Naval Science & Technology http://www.knst.kr/
Comments