Closed doannam020293 closed 6 years ago
I want to predict POS, but my data have syllable with wrong capital letters. So result is not good. So I want to retrain model without "capial word" feature, or add more lower case data set in training dataset. What do you thinks?
I think it's a possible solution, worth trying on your data. Good luck with that!
I want to predict POS, but my data have syllable with wrong capital letters. So result is not good. So I want to retrain model without "capial word" feature, or add more lower case data set in training dataset. What do you thinks?