I know what you did last summer
March 15, 2010
So I’m working away here on my little teach yourself datamining in 21 days project. One thing that is distracting me is that I’m becoming aware of the possibilities of datamining not just games but RL™ itself.
The rise of the social web is making an enormous amount of data about the average person in the street available to be mined. We seem to be at some tipping point just now, where serious work is starting to be done on this (leaving aside the question of whatever serious datamining work has been done by various shadowy arms of government).
And, for better or worse, the data could be a lot more revealing than the sort of insights your supermarket might be collecting about you. They just pay people to figure out whether you’re more likely to buy beer or soft drink with your instant noodles. The stuff that’s potentially available from Twitter, Facebook and its like, and from the various tentacles of the googlepus makes that seem like small change indeed.
One of my favourite sites at the moment is floatingsheep. Highly worth a look if you’re a dataphile at heart.
You can see from that site that concerns about privacy can be greatly exaggerated. Datamining is more interested in social trends than what the individual might be up to. The same principle applies here: my reports deal with averages – what 100K gnomes do with their time, rather than what any single gnome might be doing – but it would be easy to drill down much lower if I felt it was worth it. If I collected the achievements data and tracked specific guilds or toons to-ing and fro-ing across Azeroth, the analogy would be even clearer.
There must be a point where people beyond the alfoil hat brigade start to get a bit upset by some of the implications of social datamining. Don’t forget that Blizz had to remove all the monetary data from the armoury statistics – and there’s no sign of that data coming back – and that’s just about a game. We saw something of what might happen in the fuss over Google Buzz. How long before we see more general concerns surface out there in Web-2.0-land?