본문 바로가기
  • Home

Design and implementation of a fuzzy hash-based insider information leakage prevention system for small and medium-sized enterprises

  • Journal of The Korea Society of Computer and Information
  • Abbr : JKSCI
  • 2025, 30(10), pp.135~142
  • Publisher : The Korean Society Of Computer And Information
  • Research Area : Engineering > Computer Science
  • Received : July 22, 2025
  • Accepted : September 25, 2025
  • Published : October 31, 2025

Dae-Won Kim 1 Myung-Ho Kim 1

1숭실대학교

Accredited

ABSTRACT

With the recent increase in document leakage incidents caused by insiders in the SME environment, there is a growing need for content-based detection technology that can complement the limitations of policy-based security systems. This paper proposes a lightweight detection system that quantitatively evaluates document modification using fuzzy hash-based similarity analysis and detects the possibility of insider leakage in real time. The system is designed to calculate similarity scores for each document by parallel application of the ssdeep, sdhash, and tlsh algorithms, and determine confidentiality based on the maximum value. Experiments were conducted using a total of 1,100 text documents and modified data with various modification rates (10–100%). The results showed that sdhash achieved the best performance across all document modification rate ranges, with an ROC-AUC of 0.99947, a precision of 0.94972, a recall of 0.92016, and an F1-score of 0.93471. These results suggest that the system is effective for real-world security responses in small and medium-sized enterprise environments, and has the potential to be applied to real-time information leakage prevention systems.

Citation status

* References for papers published after 2024 are currently being built.