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

We consider the explanation problem of Graph Neural Networks (GNNs). Mostexisting GNN explanation methods identify the most important edges or nodes butfail to consider substructures, which are more important for graph data. Theonly method that considers subgraphs tries to search all possible subgraphs andidentify the most significant subgraphs. However, the subgraphs identified maynot be recurrent or statistically important. In this work, we propose a novelmethod, known as MotifExplainer, to explain GNNs by identifying importantmotifs, recurrent and statistically significant patterns in graphs. Ourproposed motif-based methods can provide better human-understandableexplanations than methods based on nodes, edges, and regular subgraphs. Givenan input graph and a pre-trained GNN model, our method first extracts motifs inthe graph using well-designed motif extraction rules. Then we generate motifembedding by feeding motifs into the pre-trained GNN. Finally, we employ anattention-based method to identify the most influential motifs as explanationsfor the final prediction results. The empirical studies on both synthetic andreal-world datasets demonstrate the effectiveness of our method.

MotifExplainer: a Motif-based Graph Neural Network Explainer