Graph Neural Networks for Misinformation Detection: Performance-Efficiency Trade-offs

2026-04-09

AI Summary

This study addresses the growing complexity of online misinformation detection models, which often incurs high computational costs and deployment challenges. Under a unified TF-IDF feature framework, it presents the first systematic comparison between lightweight graph neural networks—specifically GCN, GraphSAGE, GAT, and ChebNet—and classical machine learning approaches, including logistic regression, SVM, and MLP, in terms of both performance and inference efficiency. Evaluations across seven real-world multilingual datasets demonstrate that lightweight GNNs significantly outperform non-graph models while maintaining low inference latency. Notably, GraphSAGE achieves F1 scores of 96.8% on Kaggle and 91.9% on WELFake. These findings challenge the prevailing reliance on highly complex architectures, showing that efficient graph-based models can deliver state-of-the-art detection performance.

Abstract

The rapid spread of online misinformation has led to increasingly complex detection models, including large language models and hybrid architectures. However, their computational cost and deployment limitations raise concerns about practical applicability. In this work, we benchmark graph neural networks (GNNs) against non-graph-based machine learning methods under controlled and comparable conditions. We evaluate lightweight GNN architectures (GCN, GraphSAGE, GAT, ChebNet) against Logistic Regression, Support Vector Machines, and Multilayer Perceptrons across seven public datasets in English, Indonesian, and Polish. All models use identical TF-IDF features to isolate the impact of relational structure. Performance is measured using F1 score, with inference time reported to assess efficiency. GNNs consistently outperform non-graph baselines across all datasets. For example, GraphSAGE achieves 96.8% F1 on Kaggle and 91.9% on WELFake, compared to 73.2% and 66.8% for MLP, respectively. On COVID-19, GraphSAGE reaches 90.5% F1 vs. 74.9%, while ChebNet attains 79.1% vs. 66.4% on FakeNewsNet. These gains are achieved with comparable or lower inference times. Overall, the results show that classic GNNs remain effective and efficient, challenging the need for increasingly complex architectures in misinformation detection.

Problem

• misinformation detection
• graph neural networks
• computational efficiency
• model complexity
• practical applicability

Innovation

• Graph Neural Networks
• Misinformation Detection
• Performance-Efficiency Trade-off
• Benchmarking
• Relational Structure