Just finishing up a class where I have been doing a lot of social media mining and analytics and I thought it would be a fun given all the banter about #fakenews to do my own POTUS sentiment analysis.
- Open twitter stream with python and tweepy and grab tweets with #trump & #potus hashtags
- Sentiment analysis with textblob
- Index tweets in elasticsearch
- Create visualizations using kibana
Note: The above processes are running and the visualizations below are live. Refresh page to refresh data.
Disclaimer: This is real data pulled from the twitter API. I can assure you there is zero motivation or interest here beyond the raw data, the sheer volume of data given the topic, and of course I realize this is relevant to many others which is what makes it interesting.