Begin (1) for each [l.sub.i,j] [member of] m //m is the number of links in G (2) L([l.sub.i,j]) [left arrow] node label with higher degree in the two nodes connected by [l.sub.i,j] (3) T([l.sub.i,j]) [left arrow] 0 // initialized label modification marker of links (4) end for (5) labelList
[left arrow] the set of nonredundant labels ranged by the number of links with the same label in descending order (6) for each i [member of] labelList
(7) dnEdges [right arrow] select a set from the links whose labels are i in the network (8) for each [l.sub.i,j] [member of] dnEdges and T([l.sub.i,j]==0 (9) st [left arrow] search the triangle set [DELTA][l.sub.i,j] which include link [l.sub.i,j], and extract the set of nodes in [DELTA][l.sub.i,j] without nodes i and j.
The general format is <META http-equiv="PICS-Label"content='labellist
Input: trainSet, testSet, labelType Output: labelList
(1) classifier = svm_train(trainSet) (2) probabilityMatrix = classifier-probabilityEstimate(testSet) (3) for data in testSet (4) for label in labelType (5) if(probabilityMatrix[data][label] [greater than or equal to] 30%) (6) labelList[data].addLabel(label, probabilityMatrix[data] [label]) (7) end if (8) end for (9) end for (10) return labelList
= 37% The cultural creatives 5 5,448,250 The modernists 6 6,537,900 The sensible choosers 13 14,165,450 The edge 11 11,986,150 The fast fashion flirts 11 11,986,150 The labellists
12 13,075,800 The regulars 27 29,420,550 The out of fashion 16 17,434,400 Total 108,965,000 Table 1.