SYSTEM Cited by 1 source

Meta Data Classifier (ML-based)

Definition

Meta's scalable ML-based data classifier — an internal system first published in "Scalable data classification for security and privacy" (2020-07-21) — that automatically identifies sensitive data assets across Meta's data systems. The 2024-08-31 Privacy Aware Infrastructure post names it as the Step-1 input to PZM: "In addition to manual code inspection, we heavily rely on various techniques such as our scalable ML-based classifier to automatically identify data assets."

Role on this wiki

Stub system page — referenced as a dependency of PAI, not a primary subject. The 2020 classifier post has not been ingested; this page captures the use of the classifier by PZM so the PAI post's Step-1 integration is navigable.

Seen in

• sources/2024-08-31-meta-enforces-purpose-limitation-via-privacy-aware-infrastructure — named as an automated-discovery input to PZM's Step 1 (identify relevant assets).

Related

• systems/meta-policy-zone-manager — the primary consumer.
• systems/meta-policy-zones — downstream enforcement primitive.
• concepts/data-classification-tagging — adjacent concept (field-level sensitivity tagging; classifier output feeds tagging).
• concepts/data-annotation — annotation primitive the classifier helps populate.
• companies/meta