In November IBM and Hadoop announced their Big Data for Social Good Challenge. They invited developers to consider using big data sets and IBM’s Bluemix to do something good for society, saying “data speaks volumes about our collective behavior and society – so let’s use it to do something incredible!”
With that directive, it drew a variety of individuals and organizations, particularly those that wanted to work with large data sets, conduct analytics, and had a commitment to making the world a better place through such work. As this is right in our wheelhouse, we decided to enter the challenge. Today we were named the Grand Prize winners.
Before we decided which area of society we wanted to improve or which data sets to use, we formed a partnership with a team of students from Purdue’s D.A.T.A. Laboratory so that we could form a team that had both the professional acumen and the enthusiastic brainpower a challenge like this required. The team that came together consisted of our lab geeks (software engineers, data scientists, and front end developers) as well as four standout Purdue students who brought with them fresh technical brilliance and an interest in the project. Together, the team set about formulating and developing a plan for submission to the challenge.
Our submission Watch Flu Spread , works with Influenza data. Since as many as 20% of Americans are affected by the Flu each year, we believed that finding a way to track the spread of the virus and provide real time information about its likely path of spreading would be invaluable to the public.
Perscio’s CIO Brian Norris, MBA, RN-BC, FHIMSS shed light on the value of the app for the healthcare industry:
“It is critically important to know more about what is going on within a population and predict potential areas of impact…By giving even just a 5 day heads up on what might happen we can put a significant dent in this human and fiscal toll. The same approach can be applied to other high cost chronic diseases such as COPD, Asthma and Diabetes”.
The app uses a simple model of flu spread based on analyses, paired with a regression algorithm which learns over time. After conducting the analyses, we designed interactive visualizations to communicate the results of the computations. Using IBMs Node red tool and sentiment analysis to analyze tweets our team created a Flu Map as a public service that can help prevent the spread of flu by providing advanced disease prediction.
The team regarded the challenge as an opportunity to hone their tech skills while doing something good for the world. Alec McGail (Perscio software engineer and Purdue graduate) commented on his experience working on the challenge, “Sometimes in technology it’s easy to get away from what really matters. This project was really exciting in that I was able to use my programming and math skills to create something which will really help people.”
While the Perscio and Purdue team were the overall winners of the challenge, there were several impressive submissions. All of the teams who participated developed apps that demonstrated social good that can be done with big data. The entries can all be viewed at www.ibmhadoop.challengepost.com/submissions
Perscio will be donating the $20,000 grand prize to Purdue’s Department of Computer and Information Technology.