spaCy/examples/pipeline/custom_sentence_segmentation.py

62 lines
1.9 KiB
Python
Raw Normal View History

2018-12-02 06:26:26 +03:00
"""Example of adding a pipeline component to prohibit sentence boundaries
2018-08-05 14:33:52 +03:00
before certain tokens.
What we do is write to the token.is_sent_start attribute, which
takes values in {True, False, None}. The default value None allows the parser
to predict sentence segments. The value False prohibits the parser from inserting
a sentence boundary before that token. Note that fixing the sentence segmentation
should also improve the parse quality.
The specific example here is drawn from https://github.com/explosion/spaCy/issues/2627
Other versions of the model may not make the original mistake, so the specific
example might not be apt for future versions.
2019-03-16 16:15:49 +03:00
Compatible with: spaCy v2.0.0+
Last tested with: v2.1.0
2018-12-02 06:26:26 +03:00
"""
2018-08-05 14:33:52 +03:00
import plac
import spacy
2018-12-02 06:26:26 +03:00
2018-08-05 14:33:52 +03:00
def prevent_sentence_boundaries(doc):
for token in doc:
if not can_be_sentence_start(token):
token.is_sent_start = False
return doc
2018-12-02 06:26:26 +03:00
2018-08-05 14:33:52 +03:00
def can_be_sentence_start(token):
if token.i == 0:
return True
2019-03-16 16:15:49 +03:00
# We're not checking for is_title here to ignore arbitrary titlecased
# tokens within sentences
# elif token.is_title:
# return True
2018-08-05 14:33:52 +03:00
elif token.nbor(-1).is_punct:
return True
elif token.nbor(-1).is_space:
return True
else:
return False
2018-12-02 06:26:26 +03:00
2019-03-16 16:15:49 +03:00
@plac.annotations(
text=("The raw text to process", "positional", None, str),
spacy_model=("spaCy model to use (with a parser)", "option", "m", str),
)
def main(text="Been here And I'm loving it.", spacy_model="en_core_web_lg"):
print("Using spaCy model '{}'".format(spacy_model))
print("Processing text '{}'".format(text))
nlp = spacy.load(spacy_model)
doc = nlp(text)
sentences = [sent.text.strip() for sent in doc.sents]
print("Before:", sentences)
2018-12-02 06:26:26 +03:00
nlp.add_pipe(prevent_sentence_boundaries, before="parser")
2019-03-16 16:15:49 +03:00
doc = nlp(text)
sentences = [sent.text.strip() for sent in doc.sents]
print("After:", sentences)
2018-12-02 06:26:26 +03:00
if __name__ == "__main__":
2018-08-05 14:33:52 +03:00
plac.call(main)