MMKG: Multi-Modal Knowledge Graphs

Ye Liu, Hui Li, Alberto Garcia-Duran, Mathias Niepert, Daniel Onoro-Rubio +1 more
3/13/2019
cs.AIcs.CL

Official code: mniepert/mmkb (436 stars, BSD-3-Clause, last commit 2021-10-28).

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.

View on arXivView PDF

Code Implementations(1)

mniepert/mmkbOfficial100%
BSD-3-Clause

Several data modalities for KBs (visual, numerical, temporal, etc.)

43653PythonFeb 21, 20184 years agoBSD-3-Clause
freebaseknowledge-graph

Similar Papers

Modeling Relational Data with Graph Convolutional Networks

86% match1 repo8222017

Answering Visual-Relational Queries in Web-Extracted Knowledge Graphs

85% match1 repo4362017

Convolutional 2D Knowledge Graph Embeddings

85% match1 repo6902017

Graph Collaborative Attention Network for Link Prediction in Knowledge Graphs

85% match1 repo32025

Learning Complete Topology-Aware Correlations Between Relations for Inductive Link Prediction

85% match2023

ReaLitE: Enrichment of Relation Embeddings in Knowledge Graphs using Numeric Literals

85% match2025

Cite this paper

@article{liu2019mmkg,
  title  = {MMKG: Multi-Modal Knowledge Graphs},
  author = {Ye Liu and Hui Li and Alberto Garcia-Duran and Mathias Niepert and Daniel Onoro-Rubio and David S. Rosenblum},
  year   = {2019},
  eprint = {1903.05485},
  archivePrefix = {arXiv},
  url    = {https://arxiv.org/abs/1903.05485}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/1903.05485.svg)](https://codeofpaper.com/paper/1903.05485)

Discussion