A Probabilistic Framework for Knowledge Graph Data Augmentation

TL;DR

This paper introduces NNMFAug, a probabilistic data augmentation framework for knowledge graph completion that improves neural link prediction by generating diverse, scalable, and model-agnostic triples, leading to significant performance gains.

Contribution

The paper proposes a novel probabilistic framework, NNMFAug, for augmenting knowledge graph data to enhance neural link predictor training, addressing data scarcity issues.

Findings

NNMFAug improves baseline model performance on standard benchmarks.

The framework is efficient and scalable across different models and datasets.

Augmentation leads to more diverse and informative triples for training.

Abstract

We present NNMFAug, a probabilistic framework to perform data augmentation for the task of knowledge graph completion to counter the problem of data scarcity, which can enhance the learning process of neural link predictors. Our method can generate potentially diverse triples with the advantage of being efficient and scalable as well as agnostic to the choice of the link prediction model and dataset used. Experiments and analysis done on popular models and benchmarks show that NNMFAug can bring notable improvements over the baselines.