<![CDATA[This study developed a continuous authentication system based on keystroke patterns to enhance digital identity security. The problem arose because conventional authentication only verified users at the beginning of a session, making it vulnerable to session intrusion. The proposed approach utilized free-text keystroke dynamics by analyzing key press duration, inter-key latency, and typing speed as behavioral biometric features. Data were processed through normalization, sliding window segmentation, feature extraction, training of a hybrid deep learning model combining Convolutional Neural Network and Bidirectional Long Short-Term Memory, and evaluation using False Acceptance Rate, False Rejection Rate, and Equal Error Rate. The results showed that the system consistently identified legitimate users with high accuracy, fast detection response, and low error rates. It was concluded that keystroke-based continuous authentication provided an effective, adaptive, efficient, and non-intrusive security solution]]>
Copyrights © 2026
