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

This study utilizes community structures to address node degree biases inmessage-passing (MP) via learnable graph augmentations and novel graphtransformers. Recent augmentation-based methods showed that MP neural networksoften perform poorly on low-degree nodes, leading to degree biases due to alack of messages reaching low-degree nodes. Despite their success, most methodsuse heuristic or uniform random augmentations, which are non-differentiable andmay not always generate valuable edges for learning representations. In thispaper, we propose Community-aware Graph Transformers, namely CGT, to learndegree-unbiased representations based on learnable augmentations and graphtransformers by extracting within community structures. We first design alearnable graph augmentation to generate more within-community edges connectinglow-degree nodes through edge perturbation. Second, we propose an improvedself-attention to learn underlying proximity and the roles of nodes within thecommunity. Third, we propose a self-supervised learning task that could learnthe representations to preserve the global graph structure and regularize thegraph augmentations. Extensive experiments on various benchmark datasets showedCGT outperforms state-of-the-art baselines and significantly improves the nodedegree biases. The source code is available athttps://github.com/NSLab-CUK/Community-aware-Graph-Transformer.

Mitigating Degree Biases in Message Passing Mechanism by Utilizing Community Structures