Putting RDF2vec in Order

TL;DR

This paper improves RDF2vec node embeddings by incorporating order-sensitive word2vec variants, leading to significant performance gains on class-diverse entity tasks.

Contribution

It introduces the use of order-aware word2vec variants in RDF2vec, addressing a limitation of the original method.

Findings

Order-sensitive embeddings outperform original RDF2vec on class-diverse tasks

Performance gains are significant when entity classes vary

Order consideration enhances embedding quality for knowledge graphs

Abstract

The RDF2vec method for creating node embeddings on knowledge graphs is based on word2vec, which, in turn, is agnostic towards the position of context words. In this paper, we argue that this might be a shortcoming when training RDF2vec, and show that using a word2vec variant which respects order yields considerable performance gains especially on tasks where entities of different classes are involved.