HopGAT: Hop-aware Supervision Graph Attention Networks for Sparsely Labeled Graphs

TL;DR

This paper introduces HopGAT, a novel graph attention network that leverages hop-aware supervision and a simulated annealing strategy to improve node classification accuracy in sparsely labeled graphs, especially in biological networks.

Contribution

The study proposes a hop-aware attention supervision mechanism and a balanced learning strategy, advancing graph neural networks for sparse label scenarios.

Findings

HopGAT outperforms state-of-the-art models in sparse label settings.

In protein-protein interaction networks, performance loss is only 3.9% with 40% labeled data.

Supervised attention coefficients and learning strategies significantly enhance model effectiveness.

Abstract

Due to the cost of labeling nodes, classifying a node in a sparsely labeled graph while maintaining the prediction accuracy deserves attention. The key point is how the algorithm learns sufficient information from more neighbors with different hop distances. This study first proposes a hop-aware attention supervision mechanism for the node classification task. A simulated annealing learning strategy is then adopted to balance two learning tasks, node classification and the hop-aware attention coefficients, along the training timeline. Compared with state-of-the-art models, the experimental results proved the superior effectiveness of the proposed Hop-aware Supervision Graph Attention Networks (HopGAT) model. Especially, for the protein-protein interaction network, in a 40% labeled graph, the performance loss is only 3.9%, from 98.5% to 94.6%, compared to the fully labeled graph. Extensive experiments also demonstrate the effectiveness of supervised attention coefficient and learning strategies.