KBGAN: Adversarial Learning for Knowledge Graph Embeddings

TL;DR

KBGAN introduces an adversarial learning framework that enhances knowledge graph embedding models by using a generator to produce challenging negative samples, significantly improving link prediction performance on standard datasets.

Contribution

The paper proposes a novel adversarial training framework for knowledge graph embeddings, leveraging generator-discriminator dynamics to improve negative sampling and model performance.

Findings

Adversarial training improves link prediction accuracy.

KBGAN enhances existing models like TransE, TransD, DistMult, and ComplEx.

Significant performance gains on FB15k-237, WN18, and WN18RR datasets.

Abstract

We introduce KBGAN, an adversarial learning framework to improve the performances of a wide range of existing knowledge graph embedding models. Because knowledge graphs typically only contain positive facts, sampling useful negative training examples is a non-trivial task. Replacing the head or tail entity of a fact with a uniformly randomly selected entity is a conventional method for generating negative facts, but the majority of the generated negative facts can be easily discriminated from positive facts, and will contribute little towards the training. Inspired by generative adversarial networks (GANs), we use one knowledge graph embedding model as a negative sample generator to assist the training of our desired model, which acts as the discriminator in GANs. This framework is independent of the concrete form of generator and discriminator, and therefore can utilize a wide variety of knowledge graph embedding models as its building blocks. In experiments, we adversarially train two translation-based models, TransE and TransD, each with assistance from one of the two probability-based models, DistMult and ComplEx. We evaluate the performances of KBGAN on the link prediction task, using three knowledge base completion datasets: FB15k-237, WN18 and WN18RR. Experimental results show that adversarial training substantially improves the performances of target embedding models under various settings.