How is Twitter feeling about The 2018 Midterms?

Meet the PubNub Sentiment Machine, a realtime look into how Twitter really feels

What better way to tap into how people are feeling about something than the public online soapbox, Twitter? The PubNub Sentiment Machine pipes in the Twitter firehose, analyzing 50 tweets a second, extracting a topic, and gauging the emotion of that Tweet on a scale of 🤬to 😁.

And right now, we’re looking at the 2018 Midterms. Enjoy!

What’s Happening Under the Hood?

  • The PubNub Twitter Firehose pipes in 50 Tweets per second and Tweets are filtered by keywords.
  • Tweets that match a keyword are processed through the Amazon Comprehend Block, which serverlessly gauges the sentiment of the Tweet.
  • Results are sent in realtime to the live-visualization dashboard, powered by Initial State.
  • The entire process takes under a second and all runs on the PubNub Data Stream Network.

Want to Build One Yourself?

Share this on facebookShare this on TwitterShare this on Linkedin