Question

Your organization wants to automatically detect internal contract documents that do not match any built-in sensitive information type or standard form structure. A compliance analyst creates a custom trainable classifier and finishes training it with positive and negative sample content. After training completes, the classifier's status shows that it is ready for a validation step before it can be used in policies. What must the analyst do next before the trainable classifier can be selected in an auto-labeling policy?

Accepted Answer

Test the classifier against additional sample items and confirm the prediction accuracy, then publish it

Answer

Publish the classifier immediately, since training completion automatically makes it available to policies

Answer

Upload an EDM schema file to associate the classifier with an exact data match table

Answer

Convert the classifier into a custom sensitive information type using a keyword dictionary