Multimodal data-driven knowledge graph reasoning model for historical and cultural resources
A multi-modal data-driven knowledge graph reasoning model integrating unified encoding, graph representation learning, and rule constraints is proposed, indicating better relation completion accuracy, reasoning stability, and inference efficiency in multi-modal knowledge graph reasoning tasks.