How does unstructured data amplify re-identification risks specifically?

Through implicit identifiers in text, metadata leakage (like EXIF), and contextual re-identification from combining innocuous details.

• Implicit identifiers in text: natural language references people countless ways — full names, nicknames, job titles, family relationships, unique life events — that together pinpoint identity even with no formal "name field."
• Metadata leakage: images carry EXIF data (GPS location, device, timestamp); videos embed camera models/settings; documents store author info and edit histories — all silent privacy leaks.
• Contextual re-identification: even without explicit names, combining seemingly harmless details (location patterns, behavioral traits, social connections) can uniquely identify someone.

Tip: Before publishing a "clean" photo or document, strip metadata — the GPS tag in an EXIF header has outed more people than the image content ever did.