snopes_tweet

use GetOldTweets_modified(work in python3.6) to crawl tweets according to keywords extracted by Google Cloud API or by human label

Dependency

python3.6

pip install -r requirements.txt

Usage

run in commmand line:

run_v2.sh number

"number" is the the integer of json files' number in JsonFile directory you want to use. Just set it to be the max number of Json file number like 1200. Shell can run correctly.

You can view results of Tweet in csv format in CsvResult Dir.

Description

For a single run, I use the top 5 keywords in json which are collected by Google API and splice them into a query used by GetOldTweets. This is implemented in bash shell in run.sh (update: keywords from Googl API does not work well, so we use keywords human labelled instead.)

For multiple run, I write run_v2.sh for convenience.

Count

If you want to count the meaningful data which has some relevant tweets, just run count.sh 2000 You can view the result at Meaningful.txt.

Other detail