Yoon Kyeongchan, Park Woongjin, An Myoungsik, Jang Junho, Kim Hyukang, Journal of Internet Computing and Services, Vol. 26, No. 2, pp. 85-95, Apr. 2025
10.7472/jksii.2025.26.2.85, Full Text:  HTML
Keywords: Cyber Security, cyber assets identification, AI, Zmap, Zgrab, Asset Information, Inference

Abstract

With the advancement of digital and information communication technologies, cyberspace has emerged as a critical domain, accompanied by an exponential rise in cyber threats. Network asset identification and inference play a pivotal role in cybersecurity, particularly with the integration of artificial intelligence (AI) technologies. This study constructs a virtual network environment that simulates real-world conditions to investigate the inference of protocols, applications, and application versions. Using data collected through Nmap and Zmap+Zgrab, three inference algorithms were compared and evaluated. Experimental results indicate that the multilayer perceptron (MLP) model achieved the highest accuracy, demonstrating superior learning speed and resource efficiency. These findings highlight the potential of AI-driven asset inference models in cybersecurity. Future research will focus on enhancing performance through the integration of diverse datasets and further model optimizations.

Statistics
Cite this article