Configuring 'char_wb' for CountVectorsFeaturizer in RASA

Question

I have seen this configuration in default RASA chatbot, why do we need this setting for featurization, and in what specific languages? If I built a chatbot for non-English conversations, for example in Vietnamese, should I keep this default setup? What happen if I remove these lines? ` - name: "CountVectorsFeaturizer" analyzer: "char_wb" min_ngram: 1 max_ngram: 4

• name: "CountVectorsFeaturizer"`

I don't really understand, or see the impact of this featurizer in my team's Vietnamese chatbot. Hence, I believe there is no change in the F1-score of the intent classification, confidence score of that, and the accuracy of predicting entities, in terms of dialogue act measuring level.