Data Annotation, Model Debugging, and Data Work / Contextual Example Selection and Rare Pattern Coverage

How can semantic similarity measures improve match success between user input and predefined activity descriptions in conversational activity annotation?

Similar questions

Related papers