twitter-rumor-spreader

This app simulates the spread of retweets through a small network of twitter users. It's trained on real users and real data using the Naive Bayes Classification Algorithm.

Files of note

• The Naive Bayesian Classification algorithm is implemented in retweet_probabilities.py.
• The network is built via twitter_network.py
  • Actual tweets are downloaded via node.py
• The word counts used in the Naive Bayesian Classification algorithm are determined in relationship_trainer.py

Running Locally

To run locally, you'll need git, docker, and docker-compose:

git clone https://github.com/Didericis/twitter-rumor-spreader.git
cd twitter-rumor-spreader
docker-compose up

Once running, the site will be available at localhost:4000.

Live Version

A live version of the app can be found here.