Contextual based discovery
Use contextual based discovery to help discover sensitive data that does not follow fixed patterns.
Data Discovery supports using a Named Entity Recognition (NER) model to discover data such as names, organizations, nationalities, and political affiliations. Data patterns with the type Model use this feature, see pattern type in Configure Data Discovery patterns for more details.
Warning:
This feature requires a license check from the customer before it is enabled.
AI model-based discovery supports 5 patterns by default.
| Pattern | Description | Example text | Matching text |
|---|---|---|---|
| Person | The name of individual(s) | John Doe came by for a coffee. | John Doe |
| Nationality, religious or political groups (NRPs) | National, religious and political affiliations. | He was an American. | American |
| Location | Addresses or location information | He was in New York | New York |
| Date_Time | Dates and time information | He came at 9:30 today. | 9:30 today |
| Organization | Organization names | He works at ServiceNow. | ServiceNow |
Note:
AI Model-based discovery is only supports English for use with Real time anonymization(RTA) for tables and the Data Kit channel.