With the commencement of the 2016 Olympics, we feel that it is important that the world is informed and educated about the Zika virus. One of our team, Orens, has a close family member that has formerly worked for the World Health Organization. With these factors in mind, we have decided to build TrackVirus, a tool that tracks current and potential scenarios for spread of diseases and viruses.
What it does
TrackVirus allows users to see the risk of potentially contracting the Zika virus when traveling to certain cities by showing them an easy to understand HeatMap. In addition, if a user has disclosed travel history in a high risk area, a notification text will be sent to them, so that they may seek medical assistance and reassurance.
How we built it
Our software stack is a MySQL database with all the relevant geographic information, alongside attributes that quantify risk. This risk is then adjusted, either lowering slightly by time decay, or raising slightly when Tweets containing travel between two cities are scraped by the Twitter API. The risk is measured compared to a city’s total population, and is updated in real-time and shown to the user via various APIs, including Twilio and Google Maps.
Challenges we ran into
The coupling between various data corpora and the various APIs have proved to be the most difficult, as links between cities represent a large amount of data that has to be solved with exhaustive search. We viewed this hurdle as one involving both Big Data, as well as Big Processing.
Accomplishments that we're proud of
We have solved multiple problems that involve creative technical thinking in order to avoid the worst case scenario of a fully exhaustive search. The squaring of these circles not only prove our ability to execute, but also to conceptualize solutions that seem initially counter-intuitive.
What we learned
The most important skill that the team has learned is the process of taking multiple sources of data in order to create a simulation of potential future scenarios. This skill encompasses other sub-skills that deal with chaotic systems and their emergent properties, and to find the signal in these
What's next for Track Virus
Our immediate plans are to track scenarios for other viruses and diseases, past and present, so that we may learn about our models and assumptions in a more detailed perspective. After our models attain a high enough quality for production use, we plan on releasing it to the public in order that individuals and institutions alike can take action on preventing and combating diseases.