Most economic analysis tools are complicated and hard to understand. The better the analysis, the less understandable it is.
What it does
We trained a machine learning algorithm on over 30 economic indicators, to tell you whether the economy is doing good, or not good. Simple.
How we built it
Trained a Supervised Learning Algorithm to predict the probability that time period is in Recession (95% Accuracy, Reproducible Experiment on Github ) Used Metamind API to analyze the sentiment of Economic Phrases on Twitter Used the New York Times API to retrieve Business article of the Month Built Interactive Graph of Most important Economic Indicators using Rickshaw
Challenges we ran into
Data Wrangling the Complex Timeseries Data was hard, which is why we build an API on top of the original API in order to assist with things like dealing with misaligned timeseries
Accomplishments that we're proud of
Accurate Results: result of "Not Good" in months of reccesion in 2008/2009. New York Times pulls relavent Articles. API much easier than alternative. Metamind Sentiment Analysis looks good.
What we learned
How to do web development, and build a crawler for Economic data.
What's next for How is the Economy doing for Me
Automated content generation Seamless integration of third-party APIs Predictive analytics