Table 1 STM algorithm

1: V: set of nodes in Graph G

2: F(c): Transduction behavior function

3: S(v, w): arrived signal from node v to node w

4: C: the list of final clusters

5: PreClusters: the list of preliminary clusters

6: for each node pair(v, w) v, w ∈ V, v ≠ w do

7:    distance(v, w) ← the shortest path length from node v to node w

8:    set parameter c in function as F(c) as distance(v, w)

9:    signal(v, w) ← $S(v\to w)=\frac{d(v)}{{\displaystyle {\prod }_{i\in P(v,w)}d(i)}}F(c)$

10: end for

11: for each node v ∈ V do

12:       v. representative ← select the best scored node w for node v

13:       if cluster_w == null then

14:          make cluster_w

15:          cluster_w.add(v)

16:          PreClusters.add(cluster_w)

17:       else

18:          cluster_w.add(v)

19:       end if

20: end for

21: C ← Merge(PreClusters)