Automated FERPA Compliance Violation Detection Script

Design a Python script using pandas and regex that automatically scans student records and communication logs to detect potential FERPA compliance violations. The script should identify and flag instances of unauthorized personal information disclosure, generate detailed compliance reports, and create an audit trail with timestamp, violation type, and severity level. Implement machine learning classification to predict potential future compliance risks and integrate with a Django-based student management system.