Last Updated: 8 Oct 2012
Here is a summary of the suggestions received so far to the request for multi-dimensional social network datasets/corpora/collections (read the request here).
Please do not hesitate to contact me for further suggestions.
* Twitter social graph and celebrirty graph dataset(http://an.kaist.ac.kr/traces/WWW2010.html)
* Facebook social graph, wieghted random walks dataset and lastfm multigraph dataset on (http://odysseas.calit2.uci.edu/wiki/doku.php/public:online_social_networks). Also check out software and publications on the same link.
* Facebook 100 million dataset(http://it.slashdot.org/story/10/07/28/1350222/100-Million-Facebook-Pages-Leaked-On-Torrent-Site?art_pos=6). Data was shared on torrent sites (they might not be available anymore).
* Check out site like infochimps dataset(http://www.infochimps.com/datasets), if you can find some relevant dataset there
* You can pick up on co-authorship relationships in the academic world using Elsevier’s data: they have products that do the analysis for you, called SciVal.
Websites that can be used to create multi-dimensional social datasets:
* ArnetMiner web site, where friendships is defined based on the author/co-author, supervision, etc. Just acadmic relationship
* These two websites are based on emotions but the emotions are different and people can connect to each other in accordance to the different types of emotions, such as disliking something, being unhappy for something, etc.
Thanks to the following people for their suggestions and for liking the discussion: Rajesh Sharma, Tatiana Prokofyeva, Joe Roushar, Farhana Shah, Bahram Amini, Vineet Yadav, Christina Arulselvi, Jenny Delasalle.