Node Importance Estimation with Multiview Contrastive Representation Learning

Figure 1

An example of movie knowledge graph. Different colored boxes represent different types of nodes (e.g., movie and director), and different edge types represent different types of relations (e.g., wrote and starred-in). The numbers connected to the movies are the known number of votes (which can be used as a source of importance scores). The “plot” and “biography” are descriptions of the nodes.