Tuesday, March 23, 2010

No Secrets Anymore

Behram Mistree talks about his data mining and analytics experiment using Facebook while he was at MIT, transcript from CNN interview on CNN Newsroom below:

See video here. Note that the highlighted information below was removed from the video.

VELSHI: Let's talk more about how your personal information can be used to find out things about you, particularly on a social networking site. Let's bring in the co-author from a M.I.T. study, getting his Ph.D at Stanford right now. And Carter Jurnegean (ph) is from Boston. Same study. He is now a software developer.

OK, guys, you -- used your study. Just let me know if this is clear. You went on Facebook, and basically from people's information on there, figured out their sexual orientation. Is that basically it?

BEHRAM MISTREE, M.I.T. STUDY AUTHOR: Yes, sir. Absolutely. In essence, what we were trying to test is the age-old adage, birds of a feather flock together and have it apply it to the digital age. And is our research showed that based on your friends online, we could infer information about you, whether you disclosed that information about yourself or not.

VELSHI: Carter, is that particularly scientific? Because I guess you could sort of figure it out when you look at things like that or figure out where somebody is from, based on where their largest concentration of friends is, or where they went to college, even if they didn't put that on. Were your findings sort of a mystery to you?

CARTER JERNIGAN, M.I.T. STUDY AUTHOR: Well, I mean, I think what was novel about our research is that we've really shown that this birds-of-a-feather phenomenon applies in online communities, even when people don't publicly disclose information. So, if your profile had basically no information and it was very bare bones, just by seeing who you're associated with, we could still learn certain significant information about you.

VELSHI: Let's take it beyond sexual orientation. Where else could you be vulnerable about people learning things you didn't feel like you wanted to have known on Facebook, for instance?

JERNIGAN: That's a great question --

MISTREE: Go ahead.

There's emerging work that's suggesting that similar information could be found based on -- about political affiliations, religious affiliations, income bracket, those kinds of things. So, there's nothing to suggest that this work is just specific to sexual orientation. In reality, for this type of network data, you could have lots of privacy risks associated across a gamut of different traits.

VELSHI: Carter, if I were gay, I may not want to say that on Facebook, but it may not matter to me if that information was out there. If I live in Atlanta, I may not say I'm from Atlanta, it may not matter that it gets there.

Where do you draw the line as to protect information you don't want out there? Should I be on Facebook if I don't want people to know where I live, where I shop, who my friends are or what my sexual orientation is? Is it just too much information to be able to protect yourself?

JERNIGAN: Right. And it's -- become very difficult to keep all of that information private. And so as part of our research, we were very interested in raising awareness in these sort of surprising risks with online privacy. So as I said, again, a lot of people think oh, well, our research doesn't apply, because, you know, I don't have any information on my own profile.

VELSHI: Right.

JERNIGAN: But again, we were really interested in raising awareness, even if you think you have done the best job you can do to keep information private, that may not be the case.

VELSHI: Behram, there's a quote in your study from Scott Abrams, the creator of Dilbert. I just want to share this with you. It says, "Let's say that someday, technology will allow anyone to find out every possible thing about my life. I can compensate by being so uninteresting to that nobody could survive the process of snooping on me without lapsing into a coma."

Is that really going to be the solution?

MISTREE: Perhaps it is. But just so that you know, even if you're very uninteresting to the average person, you're very interesting to different companies. So, there's a real financial incentive for these -- for lots of large companies to have information about for marketing and other types of purposes. So, whether or not you're very uninteresting to the average person, you're very interesting to, say, an advertiser.

VELSHI: Carter, any conclusions here about how we should behave online?

JERNIGAN: I think, really, being aware of the fact that any actions you take online, you know, can have unintended privacy consequences. And so, again, it's this awareness that we think is really important.

VELSHI: But you're not advocating that people don't put information in, you're just saying they should know what it is.

JURNIGAN: I mean, I think it's going to be up to each person's own decision. I think disconnecting yourself entirely from technology is going to have its own consequences, just as well as connecting yourself with technology has consequences. And so, you know, full disclosure and being fully aware is definitely a very good first step.

VELSHI: Great study, guys. We were chopping off little bits of this technology, privacy issue to try and understand it better. Thanks for helping us. Behran Mistree is one of the co-authors of the study and the other Carter Jernigan, joining me from Boston. Thanks, guys.

MISTREE: Thank you.

No comments: