This commit introduces a novelty detection feature in the directory clustering system. The feature leverages the Jaccard distance metric to identify data drift in clusters of directories based on file updates.

Key changes:

      1. Added functionality to calculate Jaccard distance between different clusters. This metric measures the dissimilarity between clusters, aiding in the identification of novel or unusual patterns.

      2. Implemented a system to detect outliers and significant variations in the Jaccard distances over time. These may indicate potential security threats or vulnerabilities.

      3. Integrated novelty detection with the existing directory clustering system. This allows for the proactive monitoring of potential security threats such as ransomware attacks or exploited programs by observing changes in the structure and behavior of directory clusters.

      This novelty detection feature enhances the security monitoring capabilities of our system, allowing for early detection and response to potential threats.