I would like to use VADER sentiment analysis tool in Python's nltk library, but on a non English language. The language (Korean) uses another alphabet, so I would need to use a Korean tokenizer and a new sentiment dictionary. Is it possible to use the VADER code to train a model with a completely new sentiment lexicon? I believe GerVADER did it with the German language, and I was wondering if it was a reasonable challenge to try with a language that has another alphabet and different grammatical rules.
i have not tried yet
