That's sad man...REALLY sad

After reading this:

I found the idea interesting. But it posits one of the true risks in data mining. Finding what you are looking for in the datasets, even when something else entirely is really there.


  • Interesting, that kind of stuff is always fun, though of course flawed as you suggest.

    Twitter is probably one of the worst datasets to pick from for anything if you want to draw conclusions about the general population at least. The average Twitter user is vastly different from the average person in general, and as vastly different from the average internet person as well.

    Analyzing Facebook posts would be more interesting and telling, although of course, there's still the issue of people tending to exaggerate and be more of a drama-queen on the internet (fishing for responses, trying to get sympathy, etc.)
Sign In or Register to comment.