Program to run scraped news stories through Stanford's CoreNLP program.
The program pulls stories added to the database within the past day and that aren't currently parsed using CoreNLP. Once parsed, the parsetrees are placed back into the database. The program is currently set to proccess the first six sentences of a story.
This program makes extensive use of Brendan O'Connor's wrapper for CoreNLP. The current install comes from my (John Beieler) fork. The config file for CoreNLP makes use of the shift-reduce parser introduced in CoreNLP 3.4.
This pipeline depends on having CoreNLP 3.4 with the shift-reduce parser. Download the models like this:
wget http://nlp.stanford.edu/software/stanford-corenlp-full-2014-06-16.zip
unzip stanford-corenlp-full-2014-06-16.zip
mv stanford-corenlp-full-2014-06-16 stanford-corenlp
cd stanford-corenlp
wget http://nlp.stanford.edu/software/stanford-srparser-2014-07-01-models.jar
If errors persist, try changing the path in default_config.ini
from the
relative path ~/stanford-corenlp
to the full path (e.g.)
/home/ahalterman/stanford-corenlp
.
The default_config.ini
file has several options that can be changed,
including the MongoDB database and collection of stories to process and whether
all unparsed stories should be processed or just the stories added in the last
day.
python process.py
Up to a minute of [Errno 111] Connection refused
messages are normal during
startup.