Repo for social computing project.
The below will get all comments from the pittsburgh reddit between October 27th 2018 and November 27th 2018 (EST):
foo@bar:~$ python3 request.py pittsburgh 2018-10-27-00:00:00 2018-11-27-00:00:00Running the script classifer.py writes out the file about_shooting.txt which lists all of the reddit posts about the shooting. This is based on whether the post or any comment within the post contains the word with the stem "shoot" or "shooter". E.g. this would include "shooting", "shootings", "shooters", etc. Note that a post contains multiple comments.
Running the script comment2vector.py makes vectors for all comments in within the posts in about_shooting.txt. The vectors are written to vecs.txt with indexing by the sentence_index.txt provided.
Tasks 11-17:
- Pick cities
- Rule based Classifier
- Split by time: Be able to get the week, day, month and year of the shooting. Time stamp in Eastern Time to make it easier.
- Topic modeling