An improved version of an often quoted Internet resources for Subject/Verb/Object extraction using Spacy.
Uses Python 3.5+ and Spacy for its parser
pip install -r requirements
# use spacy to download its small model
python -m spacy download en_core_web_sm
Review the tests to see how it all works
python -m unittest discover -p "*_test.py"
from subject_verb_object_extract import findSVOs, nlp
tokens = nlp("Seated in Mission Control, Chris Kraft neared the end of a tedious Friday afternoon as he monitored a seemingly interminable ground test of the Apollo 1 spacecraft.")
svos = findSVOs(tokens)
print(svos)
outputs a list of tuples, each tuple containing the Subject,Verb,Object
[('Chris Kraft', 'neared', 'the end of a tedious Friday afternoon'), ('he', 'monitored', 'a interminable ground test of the Apollo spacecraft')]
Alternatively, run demo.py
to see its use.