Predicting Properties of Nodes via Community-Aware Features

TL;DR

This paper introduces community-aware node features that leverage network community structure to improve node classification, demonstrating their effectiveness and unique information content compared to traditional features.

Contribution

It proposes a new family of community-aware node features that are computationally efficient and provide additional information for node classification beyond existing features.

Findings

Community-aware features outperform classical features in classification tasks.

These features contain unique information not captured by traditional node embeddings.

Validation on synthetic and real networks confirms their effectiveness.

Abstract

This paper shows how information about the network's community structure can be used to define node features with high predictive power for classification tasks. To do so, we define a family of community-aware node features and investigate their properties. Those features are designed to ensure that they can be efficiently computed even for large graphs. We show that community-aware node features contain information that cannot be completely recovered by classical node features or node embeddings (both classical and structural) and bring value in node classification tasks. This is verified for various classification tasks on synthetic and real-life networks.