Joint Embedding-Based Entity Alignment for Knowledge Fusion of Power Standard Knowledge Graphs

Knowledge graphs have become essential tools for structuring and managing complex information, driving the digital transformation of power standards. Due to the heterogeneous sources and domain-specific nature of power standards, Power Standard Knowledge Graphs (PSKGs) often suffer from fragmented knowledge, semantic inconsistencies, and poor interoperability, making entity alignment a crucial yet challenging task. To address this issue, JEEA (Joint Embedding-Based Entity Alignment) is proposed as a novel approach that integrates structural, attribute, and name embeddings through a weighted fusion strategy. By simultaneously capturing topological and semantic features, JEEA significantly enhances alignment accuracy across heterogeneous PSKGs. Experimental results demonstrate that JEEA achieves good entity alignment performance in PSKGs, thereby facilitating more efficient knowledge fusion and intelligent power standard management.