based Predictive Model for Keylogging Attack Mitigation

No Thumbnail Available

Date

2023-12

Journal Title

Journal ISSN

Volume Title

Publisher

Lead City University

Abstract

In the ever-evolving landscape of cybersecurity, the relentless progression of cyber threats presents an ongoing challenge to the integrity of sensitive data and user credentials. Among these threats, keylogging malware has emerged as a particularly insidious vector, adept at covertly infiltrating systems, stealing login credentials, and exfiltrating valuable information. This research is driven by the imperative need to confront this menacing adversary. By delving into the subtle intricacies of human keystroke dynamics, we have engineered a groundbreaking and intelligent predictive model aimed at the early and reliable detection of keylogging attacks. The innovative character of this model stems from its amalgamation of two powerful techniques: adaptive neural networks and fuzzy logic inference. This research develops a Neuro-fuzzy predictive model using keystroke dynamics to reliably detect and mitigate ongoing keylogging threats. The model’s training process was conducted using a diverse dataset comprising over three hundred thousand keystroke samples, sourced from both simulated users and actual keyloggers. Impressively, baseline neural networks exhibited a detection accuracy rate of 99.1%. Building upon this solid foundation, the specialized Neuro-fuzzy model further elevated precision, achieving a remarkable 99.62% accuracy. This enhancement primarily stemmed from the model’s ability to distinguish between human and automated keystroke patterns, significantly reducing false positives. These results demonstrate that an adaptive Neuro-fuzzy model can reliably predict keylogging attacks in real-time based on anomalous keystroke dynamics before significant credentials or data are exfiltrated. The adaptive model provides a robust predictive solution to a rapidly evolving risk that continues to bypass traditional reactive defenses. Keywords: Neuro-fuzzy, Neuro-fuzzy model, keylogging, keylogging threats Total word count: 250 words

Description

Keywords

Neuro-fuzzy, Neuro-fuzzy model, keylogging, keylogging threats

Citation