Each day, humans upload more than 500 million photographs documenting every aspect of their lives. But while striking, this statistic pales in comparison to the vast quantity of information created not by humans, but about them. These data come from technologies as diverse as GPS-enabled Smartphones, wearable pedometers, and information captured in web logs and cookies.
Figure 1: Personal health devices such as Fitbit track metrics including distance walked, steps climbed, calories burned, and hours slept each night. Image credit: http://isource.com/wp-content/uploads/2013/05/FitBitDash.jpg
This generated information is big data, defined as “large, diverse, complex, longitudinal and/or distributed datasets generated form instruments, sensors, Internet transactions, email, video, click streams, and/or all other digital sources available today and in the future.” Big data brings tremendous potential for advancing scientific research. One researcher studying 35,000 schizophrenia patients demonstrated a genetic variant that eluded previous researchers working with smaller sample sizes. But big data also sharpens the potential for subtle, or even invisible, forms of discrimination. For example, algorithms determining which audiences receive offers for student loans could be so finely tuned that they target only people of a certain, gender, race, or income bracket.
Therefore, there is a critical need to identify how the benefits of big data can be embraced while preventing abuses. With this goal in mind, the Executive Office of the President requested a review of how big data will “transform the way we live and work and alter the relationships between government, citizens, businesses, and consumers.” The resulting report, Big Data: Seizing Opportunities, Preserving Values, was released last week.
After defining big data, the report reviews the Obama administration’s approach to open data and privacy by highlighting open data initiatives like Blue Button, which allows consumers to securely access their health information from a single portal. The report also explores national and international privacy frameworks, concluding that the most common risks to privacy—such as financial fraud—still involve “small data” sets. Private and public sector approaches to dealing with big data are described as complementary. The report concludes by identifying five focal points for continuing the dialogue on big data: maintaining and preserving privacy values, educating robustly and responsibly, preventing new modes of discrimination, ensuring responsible use by law enforcement, and harnessing big data as a public resource.
For more information, please find the report here.