Mining Twitter follower network of 36 million users

Re:Mining Twitter follower network of 36 million users by buza on Saturday April 02, 2011 @02:27AM (#2399378)

I'm curious about how these classifications were done. I graduated from the MIT Media Lab a few years back, and after looking through the result set titled:

"A small community of users associated with MIT Media lab detected using the algorithm after 7th iteration"

I saw a handful of MIT undergrads listed, many of which I'm certain had no affiliation with the Media lab, while nearly all of the others I do not recognize, even as an active Twitter user. I chose two random entries from this list: 'RFheargaile' and 'realduedate', and found both to be of extremely dubious authenticity.

[ Parent | on HN ]

Re:Mining Twitter follower network of 36 million users by aksbhat on Saturday April 02, 2011 @02:40AM (#2399391)

I agree they are not very high confidence classifications. Due to random nature of the algorithm, It is very hard to correctly label a community. Also since for most users explicit permission to follow is not required which leads to dubious/spam users. Since the data was collected in June 2009, it is possible that a lot of users might have blocked spam profiles from following them.

I guess I need to clarify this point.

[ Parent | on HN ]