140 M.D.
A medical checkup of your sleep schedule, and overall sentiment, based on your twitter activity
Given a twitter username, we can calculate the Poisson Probability of the user being awake or asleep at any given time of the day.
With this information, and with the predicted sentiment of the most recent tweets, we can provide a correlation between someone's sleep schedule and their mood.
We hope that this is informative to users in how they can improve their sleep and health.
Given a twitter username, we can calculate the Poisson Probability of the user being awake or asleep at any given time of the day.
With this information, and with the predicted sentiment of the most recent tweets, we can provide a correlation between someone's sleep schedule and their mood.
We hope that this is informative to users in how they can improve their sleep and health.
WHAT I DID
- Calculated the Poisson Distributions using noisy timestamp data from twitter.
- Approximated the above graph with a binary (on/off) graph to find sleep/wake time.
- Used D3.js + flask to visualize the results.
- Approximated the above graph with a binary (on/off) graph to find sleep/wake time.
- Used D3.js + flask to visualize the results.
SCREENSHOTS
