Use this task to create a new classifier that identifies data based on metadata patterns or matching rules.

Prerequisites

A classifier must reference an existing Data Class (For more information on creating or editing data classes, see Data Classes).

Steps

  1. In the left navigation pane, click Compliance > Data Classifiers.

  2. Click + Classifier.

  3. In the Details step:

    1. Enter a name for the classifier.

    2. Enter an optional description.

    3. Select the Data Class the classifier should assign when a match occurs.

  4. (Optional) Add tags using Tag Key and Tag Value, then click + Tag.

  5. Click Next.

  6. In the Framework step:

    1. Select the framework type (e.g., Regex, Data Type, List, or Path).

    2. Based on the selected framework, provide the required configuration values.

    3. For Data Type frameworks: choose a data type and optional minimum length. (Optional) Select whether to Include auto-incrementing columns when searching the database for this domain.

      1. Data Type: Select the database data type used to identify potential sensitive data.

      2. Minimum Length (Optional): Specify the minimum number of characters a value must contain to be considered a match.

      3. Include auto-incrementing columns when searching the database for this domain: Enable this to include auto-incrementing columns when searching the database for values that match this domain.

    4. For Regex frameworks:

      1. Enter one or more regular expressions.

      2. Checksum Type (Optional): Validates matched values using a checksum algorithm (for example, Luhn) to improve detection accuracy.

      3. Match Strength (Optional): Defines the confidence level required for a match to be considered valid.

      4. Case Sensitive (Optional): Determines whether matching should distinguish between uppercase and lowercase characters.

      5. Partial Match (Optional): Allows detection when the pattern appears within a larger value instead of matching the entire value.

      6. Reject Strength: Specifies the confidence level at which a value should be rejected as a valid match.

    5. For List frameworks:

      1. Click Choose File to upload a list file.

      2. Match Strength % (Optional): Defines the percentage of matching characters required for a value to be considered a match.

      3. Reject Strength: Specifies the confidence level at which a value should be rejected as a valid match.

      4. Tokenize input values: Splits input values into smaller tokens before performing matching.

      5. Tokenize Delimiter(s): Specifies the character(s) used to separate tokens when tokenizing input values.

    6. For Path frameworks:

      1. Field Value: Specifies the field name or value used for path-based matching.

      2. Match Type: Defines how the specified field or path should be matched during detection.

      3. Parent Value (Optional): Specifies the parent field or path used to refine the match context.

      4. Case Sensitive (Optional): Determines whether matching should distinguish between uppercase and lowercase characters.

      5. Partial Match (Optional): Allows detection when the pattern appears within a larger value instead of matching the entire value.

      6. Reject Strength: Specifies the confidence level at which a value should be rejected as a valid match.

  7. Click Next.

  8. In the Summary step, review the details, tags and the other details.

  9. If you need to make changes, click Back to update the previous steps. Otherwise, click Submit.

Last updated
Oct 8, 2026