Configuring Name Normalization and Entity Analysis Settings

IN THIS ARTICLE:

The Name Normalization and Entity Analysis (NN&EA) algorithm analyzes the people and organizations in a data set, identifying the various name variations and email addresses belonging to a single entity. Outputs of NN&EA include people profiles, organization profiles, normalized field outputs, auto-classification of entities by domain, and people mentions. The NN&EA algorithm can be run by itself but is also automatically included when running Email Threading, as this algorithm utilizes NN&EA to improve the accuracy of its results. For more details on how the algorithm works, please refer to the Name Normalization & Entity Analysis Details article.

WHO CAN PERFORM:

Users with the Create and manage sets permission can create and edit analytics sets. The Overlay results permission allows users to overlay results into Relativity. When the overlay permission is disabled, it disables the auto-overlay toggle and the overlay buttons.

Settings

Here’s a breakdown of the settings available when configuring the Name Normalization and Entity Analysis algorithm:

• Automatically Overlay Results: If set to Yes, when the set completes the analysis phase, it is automatically added to the overlay queue, and the results are published to the workspace. If users are concerned about the number of fields being added to the Relativity workspace, they may want to turn this option off.
• Reset: This option rolls back any changes made to the NN&EA settings, restoring them to their original default values.
• Populate Normalized Values:
  • Normalized From: Generates a new field based on the normalized sender of an email message.
  • Normalized To: Generates a new field based on the normalized recipients in the To field of an email message.
  • Normalized CC: Generates a new field based on the normalized recipients in the CC field of an email message.
  • Normalized BCC: Generates a new field based on the normalized recipients in the BCC field of an email message.
  • Normalized Recipients: Generates a new field based on a combination of normalized recipients in the To, CC, and BCC fields of an email message.
  • Normalized Email Participants (Top Level): Generates a new field based on a combination of the normalized sender and normalized recipients in the To, CC, and BCC fields of an email message for the top header of an email message only.
  • Normalized Email Participants (All): Generates a new field based on a combination of the normalized senders and normalized recipients in the To, CC, and BCC fields of the top-level email header and all subsumed lesser message headers in an email message.
• Domain Auto-Classification Rules:
  • ADD A RULE button: Domain auto-classification rules automatically apply functional designations to People Profiles and types to Organization Profiles based on email domains.
    • Domain: Enter one or more domain names separated by semicolons (e.g., domain1.com;domain2.com) for each domain, including all the information that occurs after the @ in the email address. Each auto-classification rule must contain at least one domain that matches the expected formatting.
    • Function: Select one function from the drop-down list. Each auto-classification rule must contain a function value.

    • Organization Type (optional): Select one organization type from the drop-down list. This is optional and can be left blank.

      

  • Preserve Manual Updates to Function/Type: When this option is on, any function or organization type manually applied to a profile will not be overwritten by any auto-classification rules. This option is turned on by default. When this option is off, all existing functions and organization types are overwritten by the auto-classification rules.

  • Reorder Rules: Use the grab icon to reorder the list of domains and control the trumping order of the Function assignment. This is important because a person can have more than one email address but can only have one function. The trumping order determines which function is applied when multiple rules could apply to the same person.

    

  • All Remaining Domains (optional): This rule applies a selected function (and optional organization type) to all remaining People and Organization Profiles not affected by the rules above. This rule/row cannot be moved up to change its trumping order and is left blank by default.

Edit Profiles

After People Profiles and Organization Profiles are generated by the Name Normalization and Entity Analysis process, further edits to Function and Organization Types are possible by interacting with the entity profiles directly. See the Edit People Profiles and Organization Profiles Details articles for more information.