QuestionQ504

Communications Mining

Why can labels receive bias warnings in UiPath Communications Mining even when they have 99% precision?

  • A They lack training examples.
  • B They were trained using the "Shuffle" option extensively.
  • C They were trained using the "Search" option extensively.
  • D They have low recall.
Explanation

Annotating bias can occur when labels are assigned disproportionately through text Search. Extensive Search use can make the training data unrepresentative and overfit the label concept to specific terms or phrases, so a high precision score does not prevent a bias warning.

Learn more

Community Discussion

No comments yet. Be the first to start the discussion!