I participated in the 2018 US Soccer hackathon, in which my team placed 2nd. Our project had several different sub-projects that combined to make a web app that enhanced the fan experience while following a soccer match. My focus during the hackathon was to derive insights from Tweets during a game to share with users in the app.
In this project, I scraped tweets during the Settle Reign v Chicago Red Stars game on 7/14/18. Keywords included the team names, player names, and the game hashtag. I collected the sentiment of each tweet and collected them into buckets. Below is the resulting graph that compared the sentiment during the game, to in-game events. During the game there were two serious injuries where sentiment was very negative, and then a huge jump in sentiment during the lone goal.
To import tweets and write them to a file, run: python3 twitter_import.py twitter_keys > twitter_out
To calculare the sentiment of each tweet in twitter_out and print out the sentiment values for tweets with non-zero sentiment, run: python3 sentiment.py
To group tweet sentiments into buckets, run: python3 twitter_bucket.py