PageRank over LiveJournal
At my day job, I research alternate algorithms to compute Google PageRank. I've become quite acquainted with the properties of the problem. So I decided I would put my knowledge to misuse and compute PageRank over the LiveJournal friends graph.
Over the course of three weeks, I collected the fdata.bml stats of everyone in the main undirected connected component of the graph. (That is, everyone reachable by a series of friend and friend-of links to me.) This totaled just over four million users. Then I computed PageRank and stored the results in a gigantic MySQL database.
I made a php interface to the database here.
The Ten Most Popular LiveJournal Users:
#10 colorfilter 8.26/10
#9 rainofbastards 8.26/10
#8 ohoneohone 8.29/10
#7 scottage_cheese 8.41/10
#6 pottersues 8.45/10
#5 lidi 8.54/10
#4 hipstomp 8.57/10
#3 ficbitches 8.85/10
#2 quizuniverse 9.88/10
#1 quizgalaxy 10/10
EDIT: This post took so long in the moderation queue that it misses some things I added later. For instance, here are the top 100 according to my database.
Over the course of three weeks, I collected the fdata.bml stats of everyone in the main undirected connected component of the graph. (That is, everyone reachable by a series of friend and friend-of links to me.) This totaled just over four million users. Then I computed PageRank and stored the results in a gigantic MySQL database.
I made a php interface to the database here.
The Ten Most Popular LiveJournal Users:
#10 colorfilter 8.26/10
#9 rainofbastards 8.26/10
#8 ohoneohone 8.29/10
#7 scottage_cheese 8.41/10
#6 pottersues 8.45/10
#5 lidi 8.54/10
#4 hipstomp 8.57/10
#3 ficbitches 8.85/10
#2 quizuniverse 9.88/10
#1 quizgalaxy 10/10
EDIT: This post took so long in the moderation queue that it misses some things I added later. For instance, here are the top 100 according to my database.