News: When Social Media Mining Gets It Wrong - Technology Review

10 August 2011 01:25 PM CDT


We already know that a tremendous amount of information about ourselves can be found at online social networks. From Facebook, to Twitter, to Foursquare, to Flickr, to, well, you name it.  We've been busy supplying personally relevant information, anecdotes, stories, photos, videos, etc. for the greater part of the past decade.  This article is interesting because it acknowledges that a tremendous amount of information is out there and that various people are beginning the process of mining the data in order to draw various conclusions.  It also acknowledges that most current data mining algorithms are actually quite inaccurate.  Most of the conclusions aren't accurate with respect to correctly identifying the individual, their personality traits, etc.  Yes, data mining has on occasion successfully unearthed certain scary pieces of data about a person, such as the correct social security number, but according to this post we've still got a ways to go with regards to improving the data mining algorithms for accuracy.  The space should be monitored closely, and possibly with governmental oversight, so that the data mining doesn't get out of hand.  Keep your eyes on the data miners.