Curvature and Representation Learning: Identifying Embedding Spaces for Relational Data


We consider the problem of learning representations of relational data in spaces of constant sectional curvature, i.e., Euclidean, Hyperbolic, and Spherical space. In this context, we explore how to identify a suitable embedding curvature for a given relational dataset. For this task, we investigate the use of a scalable heuristic based on local graph neighborhoods and evaluate it on classic benchmark graphs.

NeurIPS Relational Representation Learning