Feedback on Automatic Data Classification

Each time Collibra predicts data classes for a column, you get the opportunity to send feedback by accepting or rejecting the data class, or by adding a user-defined data class.

To improve future predictions, it is really important to send this feedback.

Sending feedback

Sending feedback is the act of accepting or rejecting the data classes that are predicted.

• Reject data class: The data class is removed from the column.
  The Data Classification Platform classification model no longer uses the sample data.
• Accept data class: The data class is added to the column.
  The sample data is permanently added to the Data Classification Platform classification model to improve future data class predictions.

For the Data Classification Platform, accepting a data class is more valuable than rejecting, but in general, we recommend that you always send feedback for every prediction. Without your feedback, the classification model cannot be retrained.

Creating user-defined classes

When columns cannot be classified, you can create your own data classes.

• Avoid duplications. Always check the list of proposed classes before creating a new data class.
• Avoid vague data classes.
• Avoid mixed data classes and accept the best applicable one.

The Data Classification Platform uses this new information to retrain the platform and improve the predictions in the future.