URL Clustering for Machine Learning Optimization (8.0.0)

FortiWeb 8.0.0 introduces URL clustering as part of its backend machine learning optimization. This enhancement addresses a common challenge in production environments where URLs often include dynamic path elements—such as IDs, tokens, or encoded values—that generate numerous syntactic variations of the same logical request.

Previously, these similar URLs were treated as unique entries, consuming excessive training capacity across ML modules and limiting detection effectiveness. With the new clustering logic, FortiWeb generalizes similar URLs into reusable structural patterns, allowing ML modules to learn and apply behavior models more efficiently.

This dynamic modeling ensures that the system focuses on learning structural patterns rather than surface-level variations, improving ML module efficiency and reducing noise.

No manual configuration is required. The URL Clustering module operates automatically when machine learning is enabled in the server policy.