COMETA consists of 20k English biomedical entity mentions from publicly available and anonymous health discussions on Reddit. All the mentions are expert-annotated with concepts...
The dataset used in the paper is a knowledge graph dataset, specifically the Freebase dataset, which is a large-scale knowledge graph. The dataset is used for multi-hop path...
M2ConceptBase is a large-scale multi-modal conceptual knowledge base with 152K concepts and 951K images. Each concept is associated with multiple fine-grained weighted images...