I followed this tutorial which looks at the sentiment of tweets about a particular topic (word) on twitter means I got to play with;
- Text Blob Sentiment Analysis using NLTK or Pattern libraries
- Twitter Streaming API
The speed at which containers get everything up is impressive (compared to installing everything manually…)
The tutorial word was ‘congress’ but you can muck around changing the word and see what comes up.
Here is a pie chart of the sentiment that was running for the topic ‘congress’ at that time.