Integrating Knowledge Graph embedding and pretrained Language Models in Hypercomplex Spaces

TL;DR

This paper introduces a novel hypercomplex space approach that integrates structural and textual knowledge representations from knowledge graphs and language models, improving link prediction accuracy.

Contribution

It proposes using hypercomplex algebra, specifically Dihedron and Quaternion representations, to unify multiple modalities in knowledge graph embedding and language models.

Findings

Outperforms existing models on benchmark datasets

Effectively combines structural and textual knowledge

Enhances link prediction accuracy

Abstract

Knowledge Graphs, such as Wikidata, comprise structural and textual knowledge in order to represent knowledge. For each of the two modalities dedicated approaches for graph embedding and language models learn patterns that allow for predicting novel structural knowledge. Few approaches have integrated learning and inference with both modalities and these existing ones could only partially exploit the interaction of structural and textual knowledge. In our approach, we build on existing strong representations of single modalities and we use hypercomplex algebra to represent both, (i), single-modality embedding as well as, (ii), the interaction between different modalities and their complementary means of knowledge representation. More specifically, we suggest Dihedron and Quaternion representations of 4D hypercomplex numbers to integrate four modalities namely structural knowledge graph embedding, word-level representations (e.g.\ Word2vec, Fasttext), sentence-level representations (Sentence transformer), and document-level representations (sentence transformer, Doc2vec). Our unified vector representation scores the plausibility of labelled edges via Hamilton and Dihedron products, thus modeling pairwise interactions between different modalities. Extensive experimental evaluation on standard benchmark datasets shows the superiority of our two new models using abundant textual information besides sparse structural knowledge to enhance performance in link prediction tasks.