Sentiment Analysis of @POTUS

https://github.com/miscalculation/sentiment-analysis-potus

Using machine learning, we created a sentiment analyzer that determines if tweets carry positive, neutral, or negative sentiment. In addition, we evaluated time series and identify trends for local variations to more accurately assess Trump’s usual twitter sentiment vs Obama’s.

When looking at Trump versus Obama’s tweets, we found that Obama had a more even distribution of positive and neutral sentiment of tweets, while having very little negative sentiment tweets. Trump had a disproportionately higher number of positive sentiment tweets over the other categories.