Research on Entity Label Value Assignment Method in Knowledge Graph

The lack of entity label values is one the problems faced by application Knowledge Graph. method automatically assigning still has shortcomings, such as costing more resources during training, leading to inaccurate value assignment because lacking semantics. In this paper, oriented domain-specific Graph, based on situation that initial all triples are completely unknown, an Entity Label Value Assignment Method (ELVAM) external and entropy proposed. ELVAM first constructs a Relationship Triples Cluster according relationship type, randomly extracts data from each cluster form Subset; then collects extended semantic text entities in subset obtain nouns. Information Entropy Conditional nouns calculated through Ontology Category Hierarchy so with moderate granularity. Finally, Pattern summarized, corresponding assigned pattern. experimental results verify effectiveness