Set up smart tags in eDiscovery (Premium)

Machine learning (ML) capabilities in Microsoft Purview eDiscovery (Premium) can help you make the decision process more efficient when reviewing case documents in a review set. Smart tags are a way to bring the ML capabilities to where the decisions are recorded: when tagging documents during review. When you create a smart tag group, then the decisions that are the result of the ML model that you've associated with the smart tag group are displayed in-line with the tags in the tag group. This helps see the ML results information in-line when you're reviewing specific documents.


If you're not an E5 customer, use the 90-day Microsoft Purview solutions trial to explore how additional Purview capabilities can help your organization manage data security and compliance needs. Start now at the Microsoft Purview compliance portal trials hub. Learn details about signing up and trial terms.

How to set up a smart tag group

  1. In a review set, select Manage review set and then select Manage tags.

  2. Select Add tag group and then select Add smart tag group.

  3. Select the ML model that you want to associate to the tag group.

    This creates a tag group and N child tags, where N is the number of possible outputs of the model. For example, the attorney-client privilege detection model has two possible outputs:

    • Positive: Use to tag documents that contain attorney-client privileged content.
    • Negative: Use to tag documents that don't contain attorney-client privileged content.

    If you select this model, a tag group with two child tags is created (one child tag named Positive and the other named Negative) for the review set. In this example, each child tag corresponds to one of the possible outputs from the attorney-client privilege detection model.

  4. Optionally, you can rename the tag group and the child tags. For example, you could rename the Positive tag to Privileged and the Negative tag to Not privileged.

How to use smart tags

When reviewing a document, the model's results are displayed next to the appropriate child tag. For example, if you have a smart tag group for attorney-client privilege detection and you review a document that is potentially privileged, the reason for that conclusion is displayed next to the appropriate tag. It's important to note that the tag isn't automatically applied to the document. The reviewer makes the decision about how to tag the document.