Abstract:The growth in social media has exacerbated the threat of fake news to individuals and communities. This draws increasing attention to developing efficient and timely rumor detection methods. The prevailing approaches resort to graph neural networks (GNNs) to exploit the post-propagation patterns of the rumor-spreading process. However, these methods lack inherent interpretation of rumor detection due to the black-box nature of GNNs. Moreover, these methods suffer from less robust results as they employ all the propagation patterns for rumor detection. In this paper, we address the above issues with the proposed Diverse Counterfactual Evidence framework for Rumor Detection (DCE-RD). Our intuition is to exploit the diverse counterfactual evidence of an event graph to serve as multi-view interpretations, which are further aggregated for robust rumor detection results. Specifically, our method first designs a subgraph generation strategy to efficiently generate different subgraphs of the event graph. We constrain the removal of these subgraphs to cause the change in rumor detection results. Thus, these subgraphs naturally serve as counterfactual evidence for rumor detection. To achieve multi-view interpretation, we design a diversity loss inspired by Determinantal Point Processes (DPP) to encourage diversity among the counterfactual evidence. A GNN-based rumor detection model further aggregates the diverse counterfactual evidence discovered by the proposed DCE-RD to achieve interpretable and robust rumor detection results. Extensive experiments on two real-world datasets show the superior performance of our method. Our code is available at <a class="link-external link-https" href="https://github.com/Vicinity111/DCE-RD" rel="external noopener nofollow">this https URL</a>.

A Novel and High-Accuracy Rumor Detection Approach Using Kernel Subtree and Deep Learning Networks

Interpretable Graph Neural Network for Social Media Rumor Detection

Detecting the Rumor Patterns Integrating Features of User, Content, and the Spreading Structure.

Rumor Detection on Social Media with Bi-Directional Graph Convolutional Networks

A Novel Fine-Grained Rumor Detection Algorithm with Attention Mechanism

Rumor detection on social media using hierarchically aggregated feature via graph neural networks

Region-enhanced Deep Graph Convolutional Networks for Rumor Detection

Call Attention to Rumors: Deep Attention Based Recurrent Neural Networks for Early Rumor Detection

An End-to-End Rumor Detection Model Based on Feature Aggregation

Rumor Detection with a novel graph neural network approach

DDGCN: Dual Dynamic Graph Convolutional Networks for Rumor Detection on Social Media

Epidemiology-informed Network for Robust Rumor Detection

Research Status of Deep Learning Methods for Rumor Detection

Rumor Detection Based on Knowledge Enhancement and Graph Attention Network

A C-GRU Neural Network for Rumors Detection

A Bi-GRU-DSA-based social network rumor detection approach

TSNN: A Topic and Structure Aware Neural Network for Rumor Detection

Rumor Detection Based on Knowledge Graph and Graph Attention Network

Explainable Rumor Detection using Inter and Intra-feature Attention Networks

Joint learning of structural and textual information on propagation network by graph attention networks for rumor detection

Rumor Detection with Diverse Counterfactual Evidence