CAREA: Cotraining Attribute and Relation Embeddings for Cross-Lingual Entity Alignment in Knowledge Graphs

Algorithm 1

Cotraining of CAREA.

	Input: Two KGs to be aligned and , a set of prealigned entity pairs, and the parameters of threshold function.
	Output: The parameters of and .
(1)	Initial iteration number ;
(2)	repeat
(3)	Reinitialize and ;
(4)	Train based on ;
(5)	;
(6)
(7)	;
(8)
(9)	Train based on ;
(10)	;
(11)	;
(12)	;
(13)
(14)
(15)	until;