Why Intel thinks data democratization is a good bet

Intel (s intc) doesn’t see money in the blending of personal data with publicly available data sets, not yet, anyway. But the chip maker does suspect that a market could form around that sort of behavior down the line. That’s why the company has been backing several initiatives around data lately, from wethedata.org to the upcoming National Day of Civic Hacking. It also introduced its own version of Hadoop. Much of these projects are targeted at democratizing data, a major goal of which is to make data easily accessible for large swaths of people.

Brandon Barnett

Brandon Barnett

Perhaps it’s not surprising that a big company like Intel retains ethnographers, anthropologists and other social scientists on staff. Those people, who work inside Intel’s Interaction and Experience Research Lab, study how people around the world use technology and how technology affects cultures.

Lately, the researchers have been exploring the implications of the changing role of data in the computing ecosystem. The idea is to figure out how the growing openness of data from companies and governments could present business opportunities for Intel, said Brandon Barnett, director of business innovation at Intel.

It’s only natural that this is a bit amorphous — Intel doesn’t know exactly what will come out of it all. But as it throws support behind data-democratization efforts, the company has in mind a few use cases that hint at where commercialization could come in to play.

  • An Intel researcher has built an application on top of multiple publicly available data sets depicting the trees and their pollen activity in Portland, Ore. The idea is for people to be able to find routes around the city that avoid trees that could cause allergy attacks. Without the application, people who move to town with allergies but no tech savvy might have a hard time commuting.
  • High school students looking at colleges should be able to plug in information about their academic performance and non-accredited extracurricular activities and get recommendations for schools they should check out, based on demographic data, grant availability and workforce demand. While the data might be out there, it’s not all navigable in one place, and it’s not interactive. (This is one area that Intel wants people to work on.)
  • Intel let Londoners look at their home energy consumption and compare it with others in the same age bracket or type of home. The trick here was to open up the data that’s relevant to certain users and to display it in a simple way. (Opower can do something related — compare people’s energy consumption with that of their neighbors.)
  • It could be possible to suggest better public-transit routes by crossing people’s transit habits with data on lateness of buses based on traffic. That sort of data can otherwise remain in a silo and never get shown to people who could actually benefit from alerts based on it.
  • A next-generation music application might be able to watch what music a person listens to, and then offer a ticket to see a preferred band when it stops by a local venue on a day the person has some free time. Such a service would take personal data and run with it by removing the hassle of identifying a band, checking out the tour schedule, finding a date that works and scoring a ticket.

It’s nice to see that Intel has ordinary people in mind, at least for now. It’s interested in constructing applications for lots of people to understand their own data and put it in context with outside data that’s narrowly tailored, “so we don’t end up with a bunch of hackers across the country coming up with better visualization tools,” Barnett said.

Of course, more public-private data mixing and analyzing could bring about the need for more computing power, and Intel wants to be ready if and when that situation arises. Currently available chips might be up for the job, or maybe a new model will be warranted. Intel has shown willingness to make custom chips for a single webscale company, Facebook, (s fb) so delivering special equipment to fit a growing use case might not be unthinkable, assuming the scale is there.

Feature image courtesy of Shutterstock user Andrea Danti.