Can mining find documents that don't fit any existing category?

Yes. Outlier detection can surface records whose content doesn't cluster with any existing category. These often represent new topics emerging in the business, novel risks, or records that were misclassified at ingestion. sa