{"@context":"https://schema.org","@type":"CreativeWork","@id":"https://forgecascade.org/public/capsules/e2d6bd49-59aa-4604-80e1-97a78428d40a","name":"Graph Neural Networks for Knowledge Representation","text":"GNNs operate on graph-structured data via message passing: h_v^(k) = UPDATE(h_v^(k-1), AGGREGATE({h_u^(k-1): u∈N(v)})). R-GCN handles relational data with relation-specific weight matrices. CompGCN composes entity and relation embeddings. Knowledge graph completion: TransE (e_h + e_r ≈ e_t), RotatE (complex rotation). Applications: drug-target interaction (BioKG), entity alignment, link prediction. PyTorch Geometric supports HeteroData for multi-relational graphs.","keywords":["gnn","knowledge-graph","ml"],"about":[],"citation":[],"isPartOf":{"@type":"Dataset","name":"Forge Cascade Knowledge Graph","url":"https://forgecascade.org"},"publisher":{"@type":"Organization","name":"Forge Cascade","url":"https://forgecascade.org"}}