Social media has significantly accelerated the rapid dissemination of information,but it also boosts propagation of fake news,posing serious challenges to public awareness and social stability.In real-world contexts,t...Social media has significantly accelerated the rapid dissemination of information,but it also boosts propagation of fake news,posing serious challenges to public awareness and social stability.In real-world contexts,the volume of trustable information far exceeds that of rumors,resulting in a class imbalance that leads models to prioritize the majority class during training.This focus diminishes the model’s ability to recognize minority class samples.Furthermore,models may experience overfitting when encountering these minority samples,further compromising their generalization capabilities.Unlike node-level classification tasks,fake news detection in social networks operates on graph-level samples,where traditional interpolation and oversampling methods struggle to effectively generate high-quality graph-level samples.This challenge complicates the identification of new instances of false information.To address this issue,this paper introduces the FHGraph(Fake News Hunting Graph)framework,which employs a generative data augmentation approach and a latent diffusion model to create graph structures that align with news communication patterns.Using the few-sample learning capabilities of large language models(LLMs),the framework generates diverse texts for minority class nodes.FHGraph comprises a hierarchical multiview graph contrastive learning module,in which two horizontal views and three vertical levels are utilized for self-supervised learning,resulting in more optimized representations.Experimental results show that FHGraph significantly outperforms state-of-the-art(SOTA)graph-level class imbalance methods and SOTA graph-level contrastive learning methods.Specifically,FHGraph has achieved a 2%increase in F1 Micro and a 2.5%increase in F1 Macro in the PHEME dataset,as well as a 3.5%improvement in F1 Micro and a 4.3%improvement in F1 Macro on RumorEval dataset.展开更多
Unsupervised learning methods such as graph contrastive learning have been used for dynamic graph represen-tation learning to eliminate the dependence of labels.However,existing studies neglect positional information ...Unsupervised learning methods such as graph contrastive learning have been used for dynamic graph represen-tation learning to eliminate the dependence of labels.However,existing studies neglect positional information when learning discrete snapshots,resulting in insufficient network topology learning.At the same time,due to the lack of appropriate data augmentation methods,it is difficult to capture the evolving patterns of the network effectively.To address the above problems,a position-aware and subgraph enhanced dynamic graph contrastive learning method is proposed for discrete-time dynamic graphs.Firstly,the global snapshot is built based on the historical snapshots to express the stable pattern of the dynamic graph,and the random walk is used to obtain the position representation by learning the positional information of the nodes.Secondly,a new data augmentation method is carried out from the perspectives of short-term changes and long-term stable structures of dynamic graphs.Specifically,subgraph sampling based on snapshots and global snapshots is used to obtain two structural augmentation views,and node structures and evolving patterns are learned by combining graph neural network,gated recurrent unit,and attention mechanism.Finally,the quality of node representation is improved by combining the contrastive learning between different structural augmentation views and between the two representations of structure and position.Experimental results on four real datasets show that the performance of the proposed method is better than the existing unsupervised methods,and it is more competitive than the supervised learning method under a semi-supervised setting.展开更多
Graph contrastive learning(GCL)has attracted extensive research interest due to its powerful ability to capture latent structural and semantic information of graphs in a self-supervised manner.Existing GCL methods com...Graph contrastive learning(GCL)has attracted extensive research interest due to its powerful ability to capture latent structural and semantic information of graphs in a self-supervised manner.Existing GCL methods commonly adopt predefined graph augmentations to generate two contrastive views.Subsequently,they design a contrastive pretext task between these views with the goal of maximizing their agreement.These methods as-sume the augmented graph can fully preserve the semantics of the original.However,typical data augmentation strategies in GCL,such as random edge dropping,may alter the properties of the original graph.As a result,previous GCL methods overlooked graph differences,potentially leading to difficulty distinguishing between graphs that are structurally similar but semantically different.Therefore,we argue that it is necessary to design a method that can quantify the dissimilarity between the original and augmented graphs to more accurately capture the relationships between samples.In this work,we propose a novel graph contrastive learning framework,named Accurate Difference-based Node-Level Graph Contrastive Learning(DNGCL),which helps the model distinguish similar graphs with slight differences by learning node-level differences between graphs.Specifically,we train the model to distinguish between original and augmented nodes via a node discriminator and employ cosine dissimilarity to accurately measure the difference between each node.Furthermore,we employ multiple types of data augmentation commonly used in current GCL methods on the original graph,aiming to learn the differences between nodes under different augmentation strategies and help the model learn richer local information.We conduct extensive experiments on six benchmark datasets and the results show that our DNGCL outperforms most state-of-the-art baselines,which strongly validates the effectiveness of our model.展开更多
基金supported by the National Key R&D Program of China(Grant No.2022YFB3104601)the Big Data Computing Center of Southeast University.
文摘Social media has significantly accelerated the rapid dissemination of information,but it also boosts propagation of fake news,posing serious challenges to public awareness and social stability.In real-world contexts,the volume of trustable information far exceeds that of rumors,resulting in a class imbalance that leads models to prioritize the majority class during training.This focus diminishes the model’s ability to recognize minority class samples.Furthermore,models may experience overfitting when encountering these minority samples,further compromising their generalization capabilities.Unlike node-level classification tasks,fake news detection in social networks operates on graph-level samples,where traditional interpolation and oversampling methods struggle to effectively generate high-quality graph-level samples.This challenge complicates the identification of new instances of false information.To address this issue,this paper introduces the FHGraph(Fake News Hunting Graph)framework,which employs a generative data augmentation approach and a latent diffusion model to create graph structures that align with news communication patterns.Using the few-sample learning capabilities of large language models(LLMs),the framework generates diverse texts for minority class nodes.FHGraph comprises a hierarchical multiview graph contrastive learning module,in which two horizontal views and three vertical levels are utilized for self-supervised learning,resulting in more optimized representations.Experimental results show that FHGraph significantly outperforms state-of-the-art(SOTA)graph-level class imbalance methods and SOTA graph-level contrastive learning methods.Specifically,FHGraph has achieved a 2%increase in F1 Micro and a 2.5%increase in F1 Macro in the PHEME dataset,as well as a 3.5%improvement in F1 Micro and a 4.3%improvement in F1 Macro on RumorEval dataset.
文摘Unsupervised learning methods such as graph contrastive learning have been used for dynamic graph represen-tation learning to eliminate the dependence of labels.However,existing studies neglect positional information when learning discrete snapshots,resulting in insufficient network topology learning.At the same time,due to the lack of appropriate data augmentation methods,it is difficult to capture the evolving patterns of the network effectively.To address the above problems,a position-aware and subgraph enhanced dynamic graph contrastive learning method is proposed for discrete-time dynamic graphs.Firstly,the global snapshot is built based on the historical snapshots to express the stable pattern of the dynamic graph,and the random walk is used to obtain the position representation by learning the positional information of the nodes.Secondly,a new data augmentation method is carried out from the perspectives of short-term changes and long-term stable structures of dynamic graphs.Specifically,subgraph sampling based on snapshots and global snapshots is used to obtain two structural augmentation views,and node structures and evolving patterns are learned by combining graph neural network,gated recurrent unit,and attention mechanism.Finally,the quality of node representation is improved by combining the contrastive learning between different structural augmentation views and between the two representations of structure and position.Experimental results on four real datasets show that the performance of the proposed method is better than the existing unsupervised methods,and it is more competitive than the supervised learning method under a semi-supervised setting.
基金supported in part by the Zhejiang Provincial Natural Science Foundation of China(LDT23F01012F01 and LDT23F01015F01)in part by the Fundamental Research Funds for the Provincial Universities of Zhejiang Grant GK229909299001-008the National Natural Science Foundation of China(62372146 and 61806061).
文摘Graph contrastive learning(GCL)has attracted extensive research interest due to its powerful ability to capture latent structural and semantic information of graphs in a self-supervised manner.Existing GCL methods commonly adopt predefined graph augmentations to generate two contrastive views.Subsequently,they design a contrastive pretext task between these views with the goal of maximizing their agreement.These methods as-sume the augmented graph can fully preserve the semantics of the original.However,typical data augmentation strategies in GCL,such as random edge dropping,may alter the properties of the original graph.As a result,previous GCL methods overlooked graph differences,potentially leading to difficulty distinguishing between graphs that are structurally similar but semantically different.Therefore,we argue that it is necessary to design a method that can quantify the dissimilarity between the original and augmented graphs to more accurately capture the relationships between samples.In this work,we propose a novel graph contrastive learning framework,named Accurate Difference-based Node-Level Graph Contrastive Learning(DNGCL),which helps the model distinguish similar graphs with slight differences by learning node-level differences between graphs.Specifically,we train the model to distinguish between original and augmented nodes via a node discriminator and employ cosine dissimilarity to accurately measure the difference between each node.Furthermore,we employ multiple types of data augmentation commonly used in current GCL methods on the original graph,aiming to learn the differences between nodes under different augmentation strategies and help the model learn richer local information.We conduct extensive experiments on six benchmark datasets and the results show that our DNGCL outperforms most state-of-the-art baselines,which strongly validates the effectiveness of our model.
