We wanted to be able to track sentiments of tweets from different states in the US, and see how it changes. If there's a major disaster or extremely positive event, it will be visible on the map.
What it does
Uses a sentiment analysis API to find positivity/negativity scores for tweets and factor them into an algorithm that determines the "moving average" sentiment for each state, which is then displayed on the frontend.
How I built it
We used the Twitter API in node.js to filter realtime tweets and sort them into their state of origin. Then, we used AYLIEN for tweet sentiment analysis, and filtered that into a moving average algorithm that gets displayed in the live frontend.
Challenges I ran into
Connecting the frontend and backend served to be more difficult than we thought, especially because most of us had never done web dev before.
Accomplishments that I'm proud of
What I learned
Learned how to use node and AmCharts.
What's next for twitter-sentiment-map
We want to be able to track more emotions in the future, and be able to highlight rapid changes in sentiment when certain events develop. We also hope to use a geocoding api to be able to more specifically highlight the locations on the map.