KNOWLEDGE GRAPH RELATIONSHIP ALIGNMENT METHOD, APPARATUS AND DEVICE, AND STORAGE MEDIUM

摘要

Provided are a knowledge graph relationship alignment method, apparatus and device, and a storage medium, wherein same relate to the field of artificial intelligence, and by means of vectoring relationships and calculating the similarity between relationship vectors, whether the relationships are similar is determined and the relationships are classified, such that the relationship alignment efficiency in a knowledge graph is improved. The method comprises: acquiring triplet data in a knowledge graph, wherein a triplet comprises a main body, a relationship and an attribute (101); converting the triplet data by means of a preset vector conversion model to obtain a relationship initial vector of each relationship (102); clustering all the relationship initial vectors by means of a preset clustering algorithm to obtain a relationship clustered vector of each relationship (103); clustering initial relationship vectors with the same relationship in the triplet data by means of the preset clustering algorithm to obtain a relationship sub-vector of each relationship (104); calculating the relationship similarity between the relationships in any two triplets on the basis of the relationship initial vectors, the relationship clustered vectors and the relationship sub-vectors (105); and classifying the relationships according to the relationship similarity (106).