English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Relational graph neural networks have garnered particular attention to encodegraph context in knowledge graphs (KGs). Although they achieved competitiveperformance on small KGs, how to efficiently and effectively utilize graphcontext for large KGs remains an open problem. To this end, we propose theRelation-based Embedding Propagation (REP) method. It is a post-processingtechnique to adapt pre-trained KG embeddings with graph context. As relationsin KGs are directional, we model the incoming head context and the outgoingtail context separately. Accordingly, we design relational context functionswith no external parameters. Besides, we use averaging to aggregate contextinformation, making REP more computation-efficient. We theoretically prove thatsuch designs can avoid information distortion during propagation. Extensiveexperiments also demonstrate that REP has significant scalability whileimproving or maintaining prediction quality. Notably, it averagely brings about10% relative improvement to triplet-based embedding methods on OGBL-WikiKG2 andtakes 5%-83% time to achieve comparable results as the state-of-the-art GC-OTE.

Simple and Effective Relation-based Embedding Propagation for Knowledge Representation Learning