On Oct 22, 2010, at 9:25 AM, Marchette, David J CIV NSWCDD, Q33 wrote: This contains a bunch of stuff. The things you really want are: ag??.edgelist classes The first two are the edge lists (each line is a pair of nodes, numbered 1-1382). The en is English (connected) and the fr is French (not connected) the edges are directed. classes is a file with one number per line indicating the class label for that node. These are: 1 -- People 2 -- Places 3 -- Dates 4 -- Things 5 -- Math things 6 -- Categories (these are special pages about categories -- a hierarchy) I don't promise the classification is perfect, but you can pretend it is. The titles files give the titles for the nodes if you want to look at them.