PII Filtering

Description

This activity scans the input text, identifies any Personally Identifiable Information (PII) or Protected Health Information (PHI), and replaces such data with a redaction marker.

Project compatibility

Windows | Cross-platform

Configuration

• Connection ID - The connection established in Integration Service. Access the dropdown menu to choose, add, or manage connections.

• Text - The document or text string containing the content to analyze for PII. This field supports String type input.
• Language - The language of the text or document input. Select a language from the available dropdown list. Defaults to English if not set. Not all PII/PHI categories are supported for all languages.

Manage Properties

Use the Manage Properties wizard to configure or use any of the object's standard or custom fields. You can select fields to add them to the activity canvas. The added standard or custom fields are available in the Properties panel (in Studio Desktop) or under Show additional properties (in Studio Web).

Additional properties

• PII/PHI category - The optional PII/PHI category to analyze the text for, such as person, person type, organization, etc. If not set, all categories are reviewed. Select the category from the available dropdown list. This field supports String[] type input.
• Minimum confidence score - The minimum confidence score required in order to be considered. This is between 0 and 1, with 0 being the lowest and 1 being the highest confidence score. If not set, all detection results are returned regardless of the confidence score. Default value is 0.75.

Output

• Redacted text - The redacted text for all PII/PHI discovered in input.
• Text analysis object - Full text analysis object.