Abstract | ||
---|---|---|
For many years, link prediction on knowledge graphs (KGs) has been a purely transductive task, not allowing for reasoning on unseen entities. Recently, increasing efforts are put into exploring semi- and fully inductive scenarios, enabling inference over unseen and emerging entities. Still, all these approaches only consider triple-based KGs, whereas their richer counterparts, hyper-relational KGs (e.g., Wikidata), have not yet been properly studied. In this work, we classify different inductive settings and study the benefits of employing hyper-relational KGs on a wide range of semi- and fully inductive link prediction tasks powered by recent advancements in graph neural networks. Our experiments on a novel set of benchmarks show that qualifiers over typed edges can lead to performance improvements of 6% of absolute gains (for the Hits@10 metric) compared to triple-only baselines. Our code is available at https://github.com/mali-git/hyper_relational_ilp. |
Year | DOI | Venue |
---|---|---|
2021 | 10.1007/978-3-030-88361-4_5 | SEMANTIC WEB - ISWC 2021 |
DocType | Volume | ISSN |
Conference | 12922 | 0302-9743 |
Citations | PageRank | References |
0 | 0.34 | 3 |
Authors | ||
7 |
Name | Order | Citations | PageRank |
---|---|---|---|
Mehdi Ali | 1 | 2 | 2.73 |
Max Berrendorf | 2 | 0 | 2.03 |
Mikhail Galkin | 3 | 1 | 2.46 |
Veronika Thost | 4 | 0 | 2.37 |
Tengfei Ma | 5 | 169 | 21.46 |
Volker Tresp | 6 | 2907 | 373.75 |
Jens Lehmann | 7 | 5375 | 355.08 |