Creates a markov chain from twitter data scraped from Trump's twitter page and outputs text imitating his writing style.
This markov chain was using 2-grams and with little tweet data. If I want to improve the realism of my model. I would increase the ngram size and scrape for data from trump's twitter. I would also improve the preprocessing of the data I scraped.
A file is also generated when this program is run that contains a count of all significant words contained in tweets.
This project contains references to API keys referenced in a file that isn't uploaded to git. If you have your own twitter api keys you can use those to test out the program.