Risk management is all about understanding the future. Big data has a role to play here.
Network World’s Mark Gibbs points to a new paper that proposes that big data be employed to predict crime. There are similarities to the movie Minority Report, except for one thing: the film showed a creepy arrangement with “precogs” in a holding tank who feed their premonitions, via brain-wave taps, to the police department to inform them of impending crimes. Rather than such a human-computer hybrid arrangement, the paper suggests computers can make all the predictions.
http://www.networkworld.com/article/2686051/big-data-business-intelligence/predicting-crime-with-big-data-welcome-to-minority-report-for-real.html
The paper, authored by a team led by Andrey Bogomolov of the University of Trento in Italy, proposes that “aggregated human behavioral data captured from the mobile network infrastructure, in combination with basic demographic information, can be used to predict crime.” They took data tracking mobile phone users’ movements available from a telecom carrier and cross-matched it against data on demographics and neighborhood conditions. The datasets reflected a select month for the London area, and was compared to actual reported instances of crime the following month. They report “an accuracy of almost 70% when predicting whether a specific area in the city will be a crime hotspot or not.”
http://arxiv.org/pdf/1409.2983v1.pdf
This has interesting implications for insurance companies. Crime prediction, of course, is a component of risk management, and such analytics could potentially be applied to property and casualty assets. But there are a lot of potential risk management applications beyond crime. Telematics data, for example, is likely to play a huge role in predicting driving behavior. Other datasets can be employed to gauge the probability and pace of wear and tear on various insured assets, from buildings to cars to boats.
Some insurers are already working on these types of problems, of course, but the ideas surfaced in the Bogomolov experiment really opens up predictive capabilities to a whole new level. Insurers, armed with highly accurate predictive data could play an increasingly proactive role in helping to prevent the calamities that lead to high claims. As the old saying goes, “an ounce of prevention is worth a pound of cure.”
