How is the best away to recognize the entity PERSONS in pt

Hello people,

I want an opinion on how to get more confidence in recognizing PEOPLE in Portuguese. I know the pipeline in Portuguese is limited, so I’m working hard to raise as much variability as possible of this entity. I am using an intention to do this training, but I am afraid that this will unbalance my other intentions. What do you recommend in this situation??

What pipeline are you using, spacy or tensorflow?

Im using spacy, and i solved this problem using the lookup who was implemented in version 0.13.3.