Supergraph Topology Feature Index for Personalized Interesting Subgraph Query in Large Labeled Graphs

Algorithm 2

Optimizational join.

	Input: Top-K_A, CM_A, Top-K_ J, CM_J, Top-K_ F, CM_F, Q
	Output: Top-K_ Result
(1)	Top-K_Result = NULL, iLQ;
(2)	int CP, O[|Partition|], CES;
(3)	O[0] ← First(Q);
(4)	O[ ] ← traverse Q;
(5)	CES ← Top-K_A; Top-K_ J; Top-K_ F;
(6)	clear(Top-K_A, Top-K_ J, Top-K_ F);
(7)	Top-K_Result ← UBV(Size-c()) for CES order by O[ ];
(8)	μ = Top-K_Result.bottom.INT;
(9)	for each Q_i ∈ Q do
(10)	;
(11)	if ∃(μi < INT(Top_i.top)) do
(12)	for each μi < INT(Top_i.top) do
(13)	for each cm in CM_i do
(14)	If INT(cm) < μi
(15)	delete cm in CM_i;
(16)	Top-K_i, CM_i ← continue UBV (Size-c()) in i partition and INT > μi;
(17)	if Top-K_i NOT NULL do
(18)	CES ← Top-K_i;
(19)	Top-K.Result ← UBV(Size-c()) for CES order by O[ ] for new matching results from i partition;
(20)	else then
(21)	return Top-K_Result;