We plan to implement a bot to compliment our Japanese customer, we are considering Rasa but wondering if it supports Japanese or not? Could anyone let me know the list of supported languages?
The tensorflow embedding pipeline technically supports any language that’s whitespace tokenized. But that’s not the case with Japanese, so I think you’d have to build a custom tokenizer there.
I was wondering the same thing if embedding would work or not for japanese since the tokenisation is not the same, however you could use FastText
Import it to spaCy and use it with Rasa
That should work i suppose
Rasa with Japanese Text. Excuse me @tamseo, Do you figure out this problems now I also trap in the same problem for Japanese Text, especially tokenization.