You plan to train your model by normalizing your utterances that you receive from your end users. This will help reduce the variance in extracting intents and entities for your LUIS application. Which API setting would you enable in your LUIS application to normalize text from Toma una taza de café to Toma una taza de cafe?

Answer options:

A.UseAllTrainingData
B.NormalizeDiacritics
C.NormalizePunctuation
D.NormalizeWordForm

Correct Answer: B
Option A is incorrect because you will use UseAllTrainingData if you would want all data as negative sample to be used for non deterministic training
Option B is correct because NormalizeDiacritics is used to replace normalizing diacritics from the utterance that would replace characters with diacritics. In this case é is converted to e.
Option C is incorrect because NormalizePunctuation would remove punctuation from the user utterance.
Option D is incorrect because NormalizeWordForm would normalize word forms.
Reference:
To learn more about normalizing user utterances, use the link given below: 

https://docs.microsoft.com/en-us/azure/cognitive-services/luis/luis-reference-application-settings