Treat Different Negatives Differently: Enriching Loss Functions with Domain and Range Constraints for Link Prediction

TL;DR

This paper proposes enriching loss functions for knowledge graph embedding models by differentiating false triples based on their semantic validity using domain and range constraints, leading to improved link prediction performance.

Contribution

It introduces a novel approach to incorporate relation signature constraints into loss functions, enhancing the quality and semantic correctness of KGEMs.

Findings

Improved MRR and Hits@10 metrics across experiments.

Better semantic correctness as measured by Sem@K.

Loss functions that differentiate negatives based on domain and range constraints.

Abstract

Knowledge graph embedding models (KGEMs) are used for various tasks related to knowledge graphs (KGs), including link prediction. They are trained with loss functions that consider batches of true and false triples. However, different kinds of false triples exist and recent works suggest that they should not be valued equally, leading to specific negative sampling procedures. In line with this recent assumption, we posit that negative triples that are semantically valid w.r.t. signatures of relations (domain and range) are high-quality negatives. Hence, we enrich the three main loss functions for link prediction such that all kinds of negatives are sampled but treated differently based on their semantic validity. In an extensive and controlled experimental setting, we show that the proposed loss functions systematically provide satisfying results which demonstrates both the generality and superiority of our proposed approach. In fact, the proposed loss functions (1) lead to better MRR and Hits@10 values, and (2) drive KGEMs towards better semantic correctness as measured by the Sem@K metric. This highlights that relation signatures globally improve KGEMs, and thus should be incorporated into loss functions. Domains and ranges of relations being largely available in schema-defined KGs, this makes our approach both beneficial and widely usable in practice.