There’s more to the biggest technology buzzword of the 21st century than stock market analysis and marketing dashboards. Data Science can now claim to be a life saver. No exaggeration here. Using massive data clustering and analysis and blending it with predictive modelling has actually helped contain an epidemic in some parts of the world, as well as give law enforcement agencies the advantage of preemptively monitoring potential crime hotspots and localities.
Here’s a closer look at these two fascinating case studies in which big data analytics showed greater promise than the sum total of its Exabyte’s:
THE EBOLA EPIDEMIC- The 2014 West Africa Ebola virus epidemic was one of the biggest epidemic outbreaks in the continent, not only in terms of the severity of the disease and the toll it took on human lives, but also in terms of the sheer speed with which it spread across countries affected by it.
Big Data, combined with geolocation platforms and information and communication technology tools empowered several agencies like the USAID and WHO to respond to the threat in near real-time and stem further social and economic disruption caused by the virus outbreak.
BACKGROUND- In 2014, the Ebola virus had already engulfed the West Africa Country of Guinea. With the epicenter of the disease rapidly expanding to include Liberia, and Sierra Leone as well, agencies had to act fast, and smart. Enter big data analytics integrated with telecom operators. A solution that proved one of the most memorable use cases of big data in disease control.
BIG DATA ANALYTICS- West Africa has some of the poorest countries in the world, with extremely low quality of infrastructure and healthcare facilities. Communication to those affected and the gathering of data from them about the epidemic were the key challenges. However, Africa has a high penetration rate of mobile phones. Approximately 89% of the population used phones with very simple features. Agencies began collaborating with telecom operators to gather metadata from phone calls, text messages and geo-location. These were then analyzed to predict the movement of the outbreak and where affected patients were moving next. Mobile Disease Control units, which were scarce, were allocated more effectively, helping to reduce disruptions caused by the outbreak, based on the data. A vast spectrum of technologies were used, ranging from Panda and Python Tools to IBM’s Watson Analytics, and even lesser known Data Visualization technologies like ESRI for overlaying the spread of the disease on maps, making it easier for relief agencies to deploy their resources effectively.
END SCENARIO- Truly the first case of frontline use of Big Data Analytics to save human lives in the middle of an outbreak, this case study has provided a template to analysts and governments alike, who are interested in tapping this exciting new technology for similar future outbreaks, globally. Malaria is being targeted next.
PREDICTIVE POLICING BY THE LOS ANGELES POLICE DEPARTMENT-
This one may seem straight out of a futuristic science fiction movie, but predictive policing isn’t around the corner, it’s happening now. Big data has made possible an unprecedented accuracy in preventing crime, and law enforcement agencies are making the most of it. Case in point? The Los Angeles Police Department.
BACKGROUND- With a population of around 3.6 million and reported violent crimes of 490.7 as of 2014, crime in Los Angeles, while not at worrying proportions, was certainly on the rise. Interestingly, in this case, the law enforcement agencies actually leveraged a mathematical model that was built to originally predict the aftershocks of an earthquake. Turns out, the probability of future crimes occurring in and around the same area of a past crime somewhat follows the same pattern. So, how did this fiction become fact?
BIG DATA ANALYTICS- The deployment of a suite of big data analytics tools from application provider Palantir was used. More than 3,500 LA police personnel were trained in the application, aptly named LASER and PredPol because of its accuracy in predicting potential crime scenes. A past crimes repository of more than 13 million crimes from the last 80 years was fed into the model.
END SCENARIO- The numbers speak for themselves! Burglaries reduced by a third (yes, a staggering 33%), a 21% reduction in violent crimes and more than 12% reduction in property crimes across the areas where the applications were implemented! The future? Implementations are already underway in New York, Nevada, Chicago, Wisconsin and many other cities!
Big data is not the future anymore. It’s the present. New Technologies and faster data management ensure better decision making, as this article showed- not just in business, but in the lives of those who have probably never even heard the term.