english

Abstrak

This study proposes an integrated framework that combines digital forensics, machine learning-based intrusion detection, and user trust assessment to enhance network security. A quantitative field-experimental approach was employed in a simulated network environment. A dataset of 5,000 network log records was used to train and evaluate a Random Forest classifier for cyberattack detection. In addition, a survey of 100 users was conducted to assess perceived security and trust using a 10-point Likert scale. The results indicate that the proposed approach improved attack detection accuracy from 80% to 95% while reducing the average detection time from 7 to 5 seconds. The Random Forest model achieved an area under the receiver operating characteristic curve (ROC-AUC) of 0.90, demonstrating strong classification performance. Furthermore, the mean user trust score increased from 6.8 to 8.5 following system implementation. These findings suggest that integrating digital forensic analysis with machine learning has the potential to improve both technical network security performance and users' perceived trust in digital systems.