Table 5.

Comparison experiment of hop number selection on SHS27k and SHS128k datasets

Datasets	Partition scheme	One-hop	One-hop $+$ two-hop	One-hop $+$ two-hop $+$ three-hop
SHS27K	Random	$88.72\pm 0.44$	89.34 $\pm 0.44$	$88.24\pm 0.95$
	BFS	$68.84\pm 5.49$	74.56 $\pm 3.03$	$68.42\pm 6.40$
	DFS	$76.92\pm 4.28$	78.20 $\pm 2.69$	$74.52\pm 4.33$
SHS148K	Random	$92.38\pm 0.08$	92.54 $\pm 0.21$	$92.28\pm 0.13$
	BFS	$70.92\pm 5.37$	73.98 $\pm 5.51$	$69.01\pm 4.36$
	DFS	$82.77\pm 1.58$	83.79 $\pm 0.95$	$81.96\pm 0.93$

Bold number represents the best value for that partition scheme