News for hackers. Stuff that matters.
Re:An open letter to Netflix from the authors of the de-anon... by randomwalker on Monday March 15, 2010 @05:11PM (#1194137)
I'm a little taken aback by the tone of some of the comments here, so I thought I'd offer a few points of clarification.

* For the longest time we mostly stuck to doing the math. We certainly didn't call for the contest to be cancelled, and we had nothing to do with the lawsuit. But when some people implied that we were responsible for the mess that ensued, we were kind of pulled into it. We posted this as a way of explaining our point of view and reaching out to see if there's a possibility of collaboration.

* The sadness we expressed is genuine. The reason we brought up Netflix's response to our paper wasn't "snark" or "gloating." Rather, we were pointing out that the cancellation of this contest was rather needless, because if they had acknowledged the privacy risks back when we published the paper, they would have had more than enough time to deploy an opt-in system for this contest. I think it is really unfortunate that that didn't happen.

* Someone wanted to know exactly what I thought of the "greater good" argument. Well, I'll tell ya. I'm vehemently opposed to it and I think it's a dangerous slippery slope. I think this point of view is enshrined in the ethos of this country -- "better let ten guilty men walk free than to convict one innocent man," etc. I don't think anyone has the moral authority to decide that the privacy concerns of a few can be sacrificed.

* There is a specific reason we chose the open letter format rather than communicating with Netflix directly. Acutally, two reasons. First, there are many data privacy researchers who are at least as qualified as we are for this role. We wanted to make sure the community had the opportunity to participate in whatever ensures, rather than just us.

Second, I'm sure there are many companies other than Netflix who have a similar need for privacy preserving data mining. If Netflix doesn't take us up, perhaps one of the others will. Bottom line, since there are multiple parties on both sides, and we don't really know who they are, we felt it is better to have this dialog in public.

* Finally, it is understandable when something like this happens to want to find someone to blame. But think twice before shooting the messenger.

Re:An open letter to Netflix from the authors of the de-anon... by houseabsolute on Tuesday March 16, 2010 @04:48AM (#1195304)
> Rather, we were pointing out that the cancellation of this contest was rather needless, because if they had acknowledged the privacy risks back when we published the paper, they would have had more than enough time to deploy an opt-in system for this contest.

And get 0.1% participation.

Re:An open letter to Netflix from the authors of the de-anon... by greendestiny on Monday March 15, 2010 @09:00PM (#1194617)
I'm not actually sure it was ethical doing this research on a set of users who had no desire to be de-anonymized. While the data was out there, and anything you came up with on other data could certainly have been used to de-anonymize the netflix data, specifically connecting the two seems like a mistake.

In fact in using this dataset you've kind of done exactly what you set out to stop - used peoples data in ways they haven't consented to.

Re:An open letter to Netflix from the authors of the de-anon... by Aron on Monday March 15, 2010 @07:23PM (#1194411)
I am gonna make a note on only a small part of this, which is that 10 guilty men is different than 'an uncountable number of guilty men'. In other words, all principals have soft margins in practice.
Re:An open letter to Netflix from the authors of the de-anon... by earl on Monday March 15, 2010 @07:30PM (#1194422)
Please. You directly enabled the lawsuit. At least have the integrity to acknowledge your part -- your pretense "Oh wow, a lawsuit just happened! But I had nothing to do with it!" -- is pretty stupid.

I'll bet money the most likely result of this lawsuits -- and your actions are a big piece of it -- is Netflix, et al, will release no more datasets to the public. Instead, only researchers under NDA will be allowed to work with the data, as was basically the tradition before this.

Good job.

Re:An open letter to Netflix from the authors of the de-anon... by jvdh on Tuesday March 16, 2010 @06:25AM (#1195389)
Please. You're acting like Netflix did nothing wrong in releasing the dataset when they knew that they could be pretty easily de-anonimized. Thus providing a privacy risk.