KG Embedding
KG: Knowledge Graph
Overview
Knowledge graph embedding studies how to encode entities and relations so models can predict missing links, answer queries, and transfer knowledge to new entities or graphs. Common approaches include bilinear and geometric scoring models, graph neural networks, temporal and hyper-relational representations, and methods that combine graph structure with text, images, or ontologies. Researchers evaluate these methods on link-prediction benchmarks and applied settings such as biomedical prediction, entity alignment, and evolving graphs. A central tension is that higher benchmark accuracy does not necessarily mean reliable or logically faithful predictions: results can depend on negative sampling, training choices, and candidate sets, while stability, privacy, and continual-learning behavior remain important challenges.
Latest papers
56 papers5 papers in the last four weeks, down 17% on the four weeks before. 0.0% of all new papers.