I am curious about whether the embedding model performs well in foreign languages

I applied embedding to the Korean data, but the classification was very different from the classification I thought.
I think that this may have been due to the difference between my classification criteria and the machine’s classification criteria, and it may have been a problem with the performance of the classifier or a problem with small data.
It’s okay if it’s not the category I want, but it’s a similar kind of sentence, so it’s not a big problem that it’s different from the category I want right now.

However, if the embedding is not trained for Korean, it is a different problem. Can you tell me how much performance (?) other languages ​​than English perform in embedding?

1 Like