Detecting network anomalies is a critical part of maintaining security and reliability within computer networks. The goal of this experiment is to investigate the multivariate data characteristics of normal, anomalous, and malicious network behaviors. By converting numerical data to categorical data, the ability to recognize attacks and generate associations becomes more efficient. By using the PVAD Algorithm to discover associations, efficiency is improved as network data, user organization safety, and threat accuracy is strengthened. Consequently, the process aims to detect the differences from profiles of normal network behaviors to enhance security threats in organizations.