How to make sure that every entity in a review gets tagged because in a review
‘SK-II immediately open eye cream large, easy-hwan bright charming eyes, you can.’.
In this review hwan is caught but SK-II didn’t come up. But SK-II was caught in other reviews!
Which recipe are you using and what are you trying to do? If you run
ner.teach, Prodigy will focus on one entity at a time and will ask you about the ones that the model is most uncertain about. This helps you select better training examples – but it’s also why you might see one example and not another.
If you already have a pre-trained model and you want to see what it actually predicts and correct that, you can use the
ner.make-gold recipe. This will stream in the model’s predictions (all that occur in the sentence) and let you correct them manually. See here for details and an example.