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

Missing node attributes is a common problem in real-world graphs. Graphneural networks have been demonstrated power in graph representation learningwhile their performance is affected by the completeness of graph information.Most of them are not specified for missing-attribute graphs and fail toleverage incomplete attribute information effectively. In this paper, wepropose an innovative node representation learning framework, Wasserstein GraphNeural Network (WGNN), to mitigate the problem. To make the most of limitedobserved attribute information and capture the uncertainty caused by missingvalues, we express nodes as low-dimensional distributions derived from thedecomposition of the attribute matrix. Furthermore, we strengthen theexpressiveness of representations by developing a novel message passing schemathat aggregates distributional information from neighbors in the Wassersteinspace. We test WGNN in node classification tasks under two missing-attributecases on both synthetic and real-world datasets. In addition, we find WGNNsuitable to recover missing values and adapt them to tackle matrix completionproblems with graphs of users and items. Experimental results on both tasksdemonstrate the superiority of our method.

Wasserstein Graph Neural Networks for Graphs with Missing Attributes