OpinionBot is a Twitter bot that uses language processing and sentiment analysis to figure out how people on Twitter feel about popular topics. Every half-hour, it visits Twitter, picks a trending topic, then analyzes the sentiment of the last 400 tweets made about that topic. It then makes posts/tweets documenting its findings. It’s written in Python and runs on a Raspberry Pi.

I wanted to create something bordering fun and useful that reflects my own struggles with understanding people. OpinionBot investigates not only the state of the sociopolitical atmosphere, but also the possibility of “fitting in” with it. One of the most interesting parts of the project is how Twitter users react toward OpinionBot’s posts, especially when its analyses are neutral or plain wrong. It's made me realize that openly stated neutrality is often the best way to get people to state their opinions, especially on a site where everyone acts as if their opinion is Very Important!

The fact that OpinionBot’s opinions are all imitations/amalgamations of other users’ is also a reflection on how our own opinions are often formed from our social environment.