Reliability of LLM Judges for Evaluating Entity Alignment.

arXiv:2610.09554v2 Announce Type: replace Abstract: Entity Alignment (EA) identifies equivalent entities across knowledge graphs and is critical for knowledge base integration and ontology merging. Evaluating EA systems at scale requires expensive expert annot