Towards Exploiting Implicit Human Feedback for Improving RDF2vec Embeddings

TL;DR

This paper enhances RDF2vec embeddings by incorporating external human feedback through Wikipedia transition probabilities, improving embedding quality in certain scenarios.

Contribution

It introduces a novel method of using external edge weights based on human feedback to guide random walks in RDF2vec, outperforming traditional approaches.

Findings

External edge weights can improve RDF2vec embeddings.

Transition probabilities from Wikipedia can serve as effective proxies for human feedback.

In some scenarios, the proposed method outperforms standard RDF2vec techniques.

Abstract

RDF2vec is a technique for creating vector space embeddings from an RDF knowledge graph, i.e., representing each entity in the graph as a vector. It first creates sequences of nodes by performing random walks on the graph. In a second step, those sequences are processed by the word2vec algorithm for creating the actual embeddings. In this paper, we explore the use of external edge weights for guiding the random walks. As edge weights, transition probabilities between pages in Wikipedia are used as a proxy for the human feedback for the importance of an edge. We show that in some scenarios, RDF2vec utilizing those transition probabilities can outperform both RDF2vec based on random walks as well as the usage of graph internal edge weights.