Accessibility Links

Benefits of Data Science for the Social Good

Social Care

Once the domain of business owners seeking insights into their customers; predictive analytics has evolved to include not only healthcare data, but social services as well. Data is changing and improving our lives and the services we use in any number of ways. The question “What if….” met with unbridled possibility. 

From Silicon Valley to the East Coast, it’s easy to imagine Data Scientists coming together to solve incredible problems. But, Nashville? It’s an unlikely hotbed of problem solvers offering degrees and jobs in the heartland.

Impact of Data Science in Social Services

Data is finding its niche in any number of unlikely places, not the least of which is addressing a social problem. In an effort to use analytics to help find solutions to society’s ills, the Data Science for Social Good (DSSG) matches recent data science graduates with data science fellows who dedicate three months to developing a solution to the problem. This includes, but is not limited to, aiding police in positive interactions with citizens, identify students in need, and helping to keep the mentally ill out of prison.

Working with nonprofits and local governments in areas as diverse as New York, Kansas and Sedosol, Mexico, data science fellows in 2016 showcased how their work was improving government services. They lead best practice workshops implementing machine learning and data-driven models at locations nationwide and beyond to improve social services. 

AIDING POLICE TO CREATE POSITIVE INTERACTIONS WITH CITIZENS

To help a police department in Nashville avoid adverse officer-citizen interactions, a team of DSSG fellows was brought in to determine best practices in allocating resources. Using past data from use of force incidents considered improper by the department as well as preventable accidents and injuries, and complaints leading to disciplinary action, these fellows develop a predictive model of behaviors. This would be used to help identify those at risk of adverse interaction with citizens.

By focusing on individual officer’s incident and complaint history combined with their characteristics, behavioral history, and work history the model helps identify high risk officers. These predictive analytics can also suggest individual intervention. Traditional blanket intervention usually involves two-thirds of the department for training. This DSSG model predictions found only five percent of Nashville officers were involved in adverse interactions.

IDENTIFIYING STUDENTS AT-RISK FOR INCARCERATION

In an effort to improve the lives of students in the Midwest, DSSG fellows developed a predictive model using education and criminal justice data. With this model, they hoped to identify students at risk for entering the juvenile justice system. The focus of the DSSG model flags less than half of the students at risk versus the traditional model.

In Tulsa, Oklahoma, the fellow’s model puts information into teacher’s hands as early as third grade. Using data from test scores, time students spent using educational software, and attendance at after-school programs, DSSG fellows built a predictive model to show teachers which students need the most help. This model ensures 95 percent of students identified receive the help they need and helps get the student on the right track.

As the movement spreads, it’s clear these data-driven efforts are widely accepted and desired. Currently offered at over 300 universities; opportunities to be part of data science for social good is growing. If you have a passion for data and analytics and want to be part of a unique opportunity, please take a look at our current vacancies or contact us to find out more. We specialize in Data and Analytics recruitment and always have a wide range of vacancies at both the junior and senior level.