header

Benchmarking Neural Embeddings for Link Prediction in Knowledge Graphs Under Semantic and Structural Changes

16 Pages Posted: 20 Jan 2021 Publication Status: Accepted

See all articles by Asan Agibetov

Asan Agibetov

Medical University of Vienna - Section for Arti cial Intelligence and Decision Support

Matthias Samwald

Section for Arti cial Intelligence and Decision Support, Medical University of Vienna

Abstract

Recently, link prediction algorithms based on neural embeddings have gained tremendous popularity in the Semantic Web community, and are extensively used for knowledge graph completion. While algorithmic advances have strongly focused on efficient ways of learning embeddings, fewer attention has been drawn to the different ways their performance and robustness can be evaluated. In this work we propose an open-source evaluation pipeline, which benchmarks the accuracy of neural embeddings in situations where knowledge graphs may experience semantic and structural changes. We define relation-centric connectivity measures that allow us to connect the link prediction capacity to the structure of the knowledge graph. Such an evaluation pipeline is especially important to simulate the accuracy of embeddings for knowledge graphs that are expected to be frequently updated.

Keywords: knowledge graphs, neural embeddings, benchmarks, evaluation, link prediction

Suggested Citation

Agibetov, Asan and Samwald, Matthias, Benchmarking Neural Embeddings for Link Prediction in Knowledge Graphs Under Semantic and Structural Changes. Journal of Web Semantics First Look , Available at SSRN: https://ssrn.com/abstract=3769876 or http://dx.doi.org/10.2139/ssrn.3769876

Asan Agibetov (Contact Author)

Medical University of Vienna - Section for Arti cial Intelligence and Decision Support ( email )

Vienna
Austria

Matthias Samwald

Section for Arti cial Intelligence and Decision Support, Medical University of Vienna

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Abstract Views
73
Downloads
3
PlumX Metrics