12:23 am June 5th, 2008 by wkallander

Someone (could not hear the name) talked about using social network analysis in the context of online auctions. Ignoring the complexity of the multiple different type of auctions that he mentioned (english v. dutch, etc), it was interesting to me, since it has elements of covert networks, particularly the game that the banned auctioneers are playing.  Since the blacklisted or fraudulent auctioneers that were removed due to seller feedback keep trying to game the system by changing identities and getting back in, and they do not want to be discovered, it is slightly familiar to me.  That they are being discovered via graph centrality measures based on the social network is downright cool.  It brings up some interesting concerns about collusive attacks in the social context, and modeling trust in social networks.

Once discovered, the culprits may then take countermeasures to hide their social context from the blocking algorithm.  One has to wonder how long the algorithm can remain effective.

And Justin made my day by mentioning that Ning supports OpenSocial - that is a very useful detail for my research.

This entry was posted on Thursday, June 5th, 2008 at 12:23 am and is filed under ECS289M. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.