Embedding Entities and Relations for Learning and Inference in Knowledge Bases

TL;DR

This paper presents a unified neural embedding framework for knowledge bases, achieving state-of-the-art link prediction results and enabling logical rule mining through relation composition.

Contribution

It introduces a simple bilinear embedding model that outperforms existing models and demonstrates how learned embeddings can be used to extract logical rules.

Findings

Bilinear embedding achieves 73.2% top-10 accuracy on Freebase.

Embeddings effectively capture relational semantics.

Relation composition via matrix multiplication aids rule mining.

Abstract

We consider learning representations of entities and relations in KBs using the neural-embedding approach. We show that most existing models, including NTN (Socher et al., 2013) and TransE (Bordes et al., 2013b), can be generalized under a unified learning framework, where entities are low-dimensional vectors learned from a neural network and relations are bilinear and/or linear mapping functions. Under this framework, we compare a variety of embedding models on the link prediction task. We show that a simple bilinear formulation achieves new state-of-the-art results for the task (achieving a top-10 accuracy of 73.2% vs. 54.7% by TransE on Freebase). Furthermore, we introduce a novel approach that utilizes the learned relation embeddings to mine logical rules such as "BornInCity(a,b) and CityInCountry(b,c) => Nationality(a,c)". We find that embeddings learned from the bilinear objective are particularly good at capturing relational semantics and that the composition of relations is characterized by matrix multiplication. More interestingly, we demonstrate that our embedding-based rule extraction approach successfully outperforms a state-of-the-art confidence-based rule mining approach in mining Horn rules that involve compositional reasoning.