Entity Alignment Method for Power Data Knowledge Graph of Semantic and Structural Information

摘要

With the continuous deepening of information construction of State Grid Corporation of China, organizing and utilizing the accumulated mass run data effectively and intelligently has become an urgent problem to solve. Knowledge graph has become an increasingly important hot technology for establishing sematic connection network for power data in full-service unified data centre. Entity alignment is one of the key steps for constructing high-quality power knowledge graph to solve the problem of a large number of entity heterogeneity and redundancy existing between different business systems. This paper proposes an entity alignment method for power data with sematic and structural information with a co-training framework. The semantic and structural models are complemented from the other after they are trained under their perspectives separately. The experiment shows the model achieves satisfactory results with higher accuracy and F1.