Predicting disease, violence, and disaster by analyzing the news
Microsoft and Technion-Israel Institute of Technology have teamed up to develop a software that can correctly predict an outcome to a natural disaster 70-90 percent of the time by mining "big data."
However, the problem with big data is just that – they’re massive dumps of data which can be really difficult to make sense of. To that end Microsoft and the Technion-Israel Institute of Technology believe they may have come up with a solution to being able to mine all that big data out there.
As we look back on the incredible damage done by things a like Hurricane Katrina and the Sandy Hook mega-storm, one has to wonder if those damages could have been lessened, or at least made us better prepared for the aftermath if we knew about them in more detail beforehand.
Computer scientists believe mining big data could lead to better outcomes to catastrophic events like hurricanes and other natural disasters. It is also the subject that Microsoft and Technion-Israel Institute of Technology researchers think that they may have found the answer for.
The team of researchers has developed software that can tap into 22 years of New York Times archives, as well as news reports from 90 other data sources on the web, and predict where disease outbreaks, violence and natural disasters might occur.
One of the examples given by the team was when the software was used on historical data from 2006 to correctly predict a cholera outbreak based on reports of drought in Angola.
So far in testing the software has been right about 70 to 90 percent of the time.
via Business Insider