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

The prevailing graph neural network models have achieved significant progressin graph representation learning. However, in this paper, we uncover anever-overlooked phenomenon: the pre-trained graph representation learning modeltested with full graphs underperforms the model tested with well-pruned graphs.This observation reveals that there exist confounders in graphs, which mayinterfere with the model learning semantic information, and current graphrepresentation learning methods have not eliminated their influence. To tacklethis issue, we propose Robust Causal Graph Representation Learning (RCGRL) tolearn robust graph representations against confounding effects. RCGRLintroduces an active approach to generate instrumental variables underunconditional moment restrictions, which empowers the graph representationlearning model to eliminate confounders, thereby capturing discriminativeinformation that is causally related to downstream predictions. We offertheorems and proofs to guarantee the theoretical effectiveness of the proposedapproach. Empirically, we conduct extensive experiments on a synthetic datasetand multiple benchmark datasets. The results demonstrate that compared withstate-of-the-art methods, RCGRL achieves better prediction performance andgeneralization ability.

Robust Causal Graph Representation Learning against Confounding Effects