MMKG: Multi-Modal Knowledge Graphs

TL;DR

MMKG is a multi-modal knowledge graph dataset containing numerical features, images, and entity alignments, designed to advance multi-modal learning and entity matching in knowledge graphs.

Contribution

This paper introduces MMKG, a novel multi-modal knowledge graph dataset with integrated images and features, enabling new research in multi-modal knowledge graph learning.

Findings

Multi-modal features improve link prediction accuracy

The dataset supports entity matching across knowledge graphs

Multi-relational learning benefits from combined feature types

Abstract

We present MMKG, a collection of three knowledge graphs that contain both numerical features and (links to) images for all entities as well as entity alignments between pairs of KGs. Therefore, multi-relational link prediction and entity matching communities can benefit from this resource. We believe this data set has the potential to facilitate the development of novel multi-modal learning approaches for knowledge graphs.We validate the utility ofMMKG in the sameAs link prediction task with an extensive set of experiments. These experiments show that the task at hand benefits from learning of multiple feature types.