Spacy recognizes gonna as an adverb

[nilehmann]

In sentences like I'm gonna send an invitation spacy recognizes gonna as an adverb. This messes up dependency parsing. Stanford instead tokenizes gonna as two tokens (gon=VBG, na=TO), so its behaves like going to. I found this behaviour more appropriate.

[honnibal]

Thanks. We'll add this to the tokenizer exceptions.

If you want a workaround immediately, you should be able to do this at run-time. See here: https://spacy.io/docs/usage/customizing-tokenizer

[ines]

Adding this on the organize-language-data branch in the new format so it'll be fixed in v2.

[lock]

This thread has been automatically locked since there has not been any recent activity after it was closed. Please open a new issue for related bugs.