Mitigating Semantic Confusion from Hostile Neighborhood for Graph Active Learning

TL;DR

This paper introduces SAG, a semantic-aware active learning framework for graphs that reduces semantic confusion and improves node classification by evaluating node influence with semantic features and maintaining diversity and class balance.

Contribution

The paper proposes a novel SAG framework that mitigates semantic confusion in graph active learning using semantic features, prototype-based criteria, and improved query policies.

Findings

SAG significantly improves node classification accuracy.

SAG outperforms previous active learning methods on benchmark graphs.

The framework's effectiveness is validated through extensive experiments and ablation studies.

Abstract

Graph Active Learning (GAL), which aims to find the most informative nodes in graphs for annotation to maximize the Graph Neural Networks (GNNs) performance, has attracted many research efforts but remains non-trivial challenges. One major challenge is that existing GAL strategies may introduce semantic confusion to the selected training set, particularly when graphs are noisy. Specifically, most existing methods assume all aggregating features to be helpful, ignoring the semantically negative effect between inter-class edges under the message-passing mechanism. In this work, we present Semantic-aware Active learning framework for Graphs (SAG) to mitigate the semantic confusion problem. Pairwise similarities and dissimilarities of nodes with semantic features are introduced to jointly evaluate the node influence. A new prototype-based criterion and query policy are also designed to maintain diversity and class balance of the selected nodes, respectively. Extensive experiments on the public benchmark graphs and a real-world financial dataset demonstrate that SAG significantly improves node classification performances and consistently outperforms previous methods. Moreover, comprehensive analysis and ablation study also verify the effectiveness of the proposed framework.