Sentence splitting

[mfilipov]

Python version: 3.4.3, latest spaCy (0.85), redownloaded all data (python -m spacy.en.download all).

It looks like there's something wrong with sentence splitting. Example:

pipeline = spacy.en.English()

tokens = pipeline("The Germans have their Faust; but Faust is a tragedy with a cosmic philosophic theme.")

root_inds = [ind for ind, token in enumerate(tokens) if token.dep_ == "ROOT"]

root_inds has two elements - corresponding sentences are "The Germans have their Faust; but" and "Faust is a tragedy with a cosmic philosophical theme.".

I guess this is related to the ROOT bug #57.

[honnibal]

Thanks — working on this.

[honnibal]

Just released version 0.86.

Your example now parses correctly, and accuracy is up on aggregate. Further improvements to sentence boundary detection accuracy should be forth-coming.

Please keep reporting prominent failures as they occur.

[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.