English Grammar Discrimination Training Network Model and Search Filtering

Algorithm 1

English grammar discriminative training network model algorithm.

(i)	Input: Any intermediate state Si; The initial state is S0
(ii)	Output: Vertex mapping pairs for target graph G1 and query subgraph G2
(iii)	Match(s)
(iv)	Begin
(v)	If (1VI(S) contains all vertices in query subgraph G2)//G2 isomorphism in GI has been found
(vi)	Output M (S);//Subgraph, end of search
(vii)	Else
(viii)	The matching point pair set P(S) is calculated according to the current local matching M(S).
(ix)	For Each p in P(S)//Traversal matching point pair set F(S)
(x)	If(in the case of p=(n,m)), the feasibility function F(S,n,m} = true)//If the matching p is added, it is feasible
(xi)	s' = s U p; Match(s');//L adds P to s and recursively calls Match(s)} to continue the search
(xii)	End For Each
(xiii)	Restore the data structure and trace back to the previous state.//No child of the isomorphism has been found after multiple calls to Match(s)
(xiv)	//Figure, which indicates that the current state cannot be expanded to be feasible
(xv)	The knife graph isomorphism matches, then goes back to the previous state
(xvi)	End