Weekend Round-Up: AOL Mistakes Edition
Good Monday, everyone. Lisa is traveling today on her intrepid mission to bring you back reports on the very best of SES San Jose. Tune in all week for her updates on sessions, parties and more. I’m manning the desk back here at home so that the rest of the news doesn’t slip through the cracks.
Okay, let’s begin!
Google Wants Their MTV
The hippest thing of the new millennium and the hippest thing of the 1980s are together at last! Google and MTV struck up a deal to syndicate MTV content on the web. WSJ has the story. (Subscription required, sorry.)
Eric Schmidt sees it as exciting and revolutionary, “Our technology takes MTV’s video, marries it to an ad, and shows it on a third site. No one’s ever done this before. If this works, it would be a very large business for all players.”
No word yet on whether or not this will resemble more the early years of MTV (you know, when they played music videos?) or the current iteration where the music is hard to find but the spoiled rich kids are plentiful.
AOL’s Privacy Blunder
People have long feared all the information that search engines gather on their users. How will they use it? Who will they give it to? Exactly how much of it is directly tied back to a specific user?
Well, the answers from AOL are apparently ‘however they like’, ‘everyone’ and unfortunately, ‘all of it.’
AOL released a data set of over 20m user queries sampled over 3 months from about 500k users. It was a great opportunity to mine their information for personalization, query refinement and many other search patterns. There was just one glitch–each of the users were identified by number which, based on their search history, made it very easy to track an ID number back to a particular user.
As Phillip explains:
Based on a sequence of searches it is often trivial to connect a person to a user ID. For example, user 500 may search for "link:mysite.com", and then user 500 may search for the name "John Doe." Now you can verify that mysite.com’s webmaster is John Doe from San Francisco, and you have a good indicator that user 500 is indeed John Doe. Finally, you look at other queries from this user – like, "jobs San Francisco" – and you have strong indicators that John Doe is looking for a job behind his current boss’s back.
Happily AOL realized the error of their ways and took down the information shortly after launch but not before Michael Arrington summed it up in the best way possible, “The utter stupidity of this is staggering.”
AOL’s official response? “This was a screw up, and we’re angry and upset about it. It was an innocent enough attempt to reach out to the academic community with new research tools, but it was obviously not appropriately vetted, and if it had been, it would have been stopped in an instant.”
Unfortunately, it’s too late. The whole dataset is still available through mirror sites. Aren’t you glad you don’t use AOL?
Google Shares Data
Speaking of giving away search data to the research community, Google gave away some of their internal numbers over the weekend. Unlike AOL, they did it right; this data can’t be tracked back to users. Thanks for the care, guys. Good to see that your ‘don’t be evil’ guideline still seems to be firmly in place.
Google’s research blog has the full scoop on what kind of information is being made available.
What would a weekend update be without mention of Matt? Coming to us from the Cutts Cave (kinda like the Batcave but without the giant penny) are a plea for his buddies in the Googleplex to stop the madness and a new video on the best way to navigate through SES San Jose.
And with that this recap has come full circle. Have a great week, everybody.