Abstract:The truth is significantly hampered by massive rumors that spread along with breaking news or popular topics. Since there is sufficient corpus gathered from the same domain for model training, existing rumor detection algorithms show promising performance on yesterday’s news. However, due to a lack of substantial training data and prior expert knowledge, they are poor at spotting rumors concerning unforeseen events, especially those propagated in different languages (i.e., low-resource regimes). In this paper, we propose a simple yet effective framework with unified contrastive transfer learning, to detect rumors by adapting the features learned from well-resourced rumor data to that of the low-resourced with only few-shot annotations. More specifically, we first represent rumor circulated on social media as an undirected topology for enhancing the interaction of user opinions, and then train the propagation-structured model via a unified contrastive paradigm to mine effective clues simultaneously from both post semantics and propagation structure. Our model explicitly breaks the barriers of the domain and/or language issues, via language alignment and a novel domain-adaptive contrastive learning mechanism. To well-generalize the representation learning using a small set of annotated target events, we reveal that rumor-indicative signal is closely correlated with the uniformity of the distribution of these events. We design a target-wise contrastive training mechanism with three event-level data augmentation strategies, capable of unifying the representations by distinguishing target events. Extensive experiments conducted on four low-resource datasets collected from real-world microblog platforms demonstrate that our framework achieves much better performance than state-of-the-art methods and exhibits a superior capacity for detecting rumors at early stages.

Cross-Domain Social Rumor-Propagation Model Based on Transfer Learning

Interpretable Graph Neural Network for Social Media Rumor Detection

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

Towards low-resource rumor detection: Unified contrastive transfer with propagation structure

A Unified Contrastive Transfer Framework with Propagation Structure for Boosting Low-Resource Rumor Detection

A Rumor & Anti-rumor Propagation Model Based on Data Enhancement and Evolutionary Game

Cross-Task Rumor Detection: Model Optimization Based on Model Transfer Learning and Graph Convolutional Neural Networks (GCNs)

Rumor2vec: A rumor detection framework with joint text and propagation structure representation learning

Region-enhanced Deep Graph Convolutional Networks for Rumor Detection

Unsupervised Rumor Detection Based on Propagation Tree VAE

Rumor Detection on Social Media with Temporal Propagation Structure Optimization

Modeling microscopic and macroscopic information diffusion for rumor detection

Rumor Detection with a novel graph neural network approach

Unifying Multimodal Interactions for Rumor Diffusion Prediction with Global Hypergraph Modeling

Neural network approaches for rumor stance detection: Simulating complex rumor propagation systems

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

Research Status of Deep Learning Methods for Rumor Detection

A group behavior prediction model based on sparse representation and complex message interactions

A Rumor Detection Model Incorporating Propagation Path Contextual Semantics and User Information

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

Modeling the Uncertainty of Information Propagation for Rumor Detection: A Neuro-Fuzzy Approach