SimplE Embedding for Link Prediction in Knowledge Graphs

Explore this paper's citation graph

Summary

It is proved SimplE is fully expressive and derive a bound on the size of its embeddings for full expressivity and shown empirically that, despite its simplicity, SimplE outperforms several state-of-the-art tensor factorization techniques.

Type
preprint
Published
2018-02-13
Cited by
869
References
69
Access
Open access

Keywords

Simple (philosophy), Embedding, Computer science, Theoretical computer science, Link (geometry)

References

Cited by

Related papers