From One Point to A Manifold: Knowledge Graph Embedding For Precise Link Prediction

TL;DR

This paper introduces ManifoldE, a manifold-based knowledge graph embedding method that enhances the precision of link prediction by expanding entity representations from points to manifolds, addressing limitations of previous models.

Contribution

The paper proposes a novel manifold-based embedding principle that improves the algebraic and geometric modeling of knowledge graphs for more accurate link prediction.

Findings

Achieves substantial improvements over state-of-the-art baselines.

Maintains high efficiency in prediction tasks.

Addresses issues of ill-posed algebraic systems and overstrict geometric forms.

Abstract

Knowledge graph embedding aims at offering a numerical knowledge representation paradigm by transforming the entities and relations into continuous vector space. However, existing methods could not characterize the knowledge graph in a fine degree to make a precise prediction. There are two reasons: being an ill-posed algebraic system and applying an overstrict geometric form. As precise prediction is critical, we propose an manifold-based embedding principle (\textbf{ManifoldE}) which could be treated as a well-posed algebraic system that expands the position of golden triples from one point in current models to a manifold in ours. Extensive experiments show that the proposed models achieve substantial improvements against the state-of-the-art baselines especially for the precise prediction task, and yet maintain high efficiency.