Abstract

 Neural models operating over structured spaces such as knowledge graphs require a continuous embedding of the discrete elements of this space (such as entities) as well as the relationships between them. Relational embeddings with high expressivity, however, have high model complexity, making them computationally difficult to train. We propose a new family of embeddings for knowledge graphs that interpolate between a method with high model complexity and one, namely Holographic embeddings (H OL E), with low dimensionality and high training efficiency. This interpolation, termed H OL E X, is achieved by concatenating several linearly perturbed copies of original H OL E. We formally characterize the number of perturbed copies needed to provably recover the full entity-entity or entity-relation interaction matrix, leveraging ideas from Haar wavelets and compressed sensing. In practice, using just a handful of Haar-based or random perturbation vectors results in a much stronger knowledge completion system. On the Freebase FB15K dataset, H OL E X outperforms originally reported H OL E by 14.7% on the HITS@10 metric, and the current path-based state-of-the-art method, PTransE, by 4% (absolute).