Rumor Detection Method Research Articles

Rumor Detection With Hierarchical Representation on Bipartite Ad Hoc Event Trees.

The rapid growth of social media has caused tremendous effects on information propagation, raising extreme challenges in detecting rumors. Existing rumor detection methods typically exploit the reposting propagation of a rumor candidate for detection by regarding all reposts to a rumor candidate as a temporal sequence and learning semantics representations of the repost sequence. However, extracting informative support from the topological structure of propagation and the influence of reposting authors for debunking rumors is crucial, which generally has not been well addressed by existing methods. In this article, we organize a claim post in circulation as an ad hoc event tree, extract event elements, and convert it into bipartite ad hoc event trees in terms of both posts and authors, i.e., author tree and post tree. Accordingly, we propose a novel rumor detection model with hierarchical representation on the bipartite ad hoc event trees called BAET. Specifically, we introduce word embedding and feature encoder for the author and post tree, respectively, and design a root-aware attention module to perform node representation. Then we adopt the tree-like RNN model to capture the structural correlations and propose a tree-aware attention module to learn tree representation for the author tree and post tree, respectively. Extensive experimental results on two public Twitter datasets demonstrate the effectiveness of BAET in exploring and exploiting the rumor propagation structure and the superior detection performance of BAET over state-of-the-art baseline methods.

MVACLNet: A Multimodal Virtual Augmentation Contrastive Learning Network for Rumor Detection

In today’s digital era, rumors spreading on social media threaten societal stability and individuals’ daily lives, especially multimodal rumors. Hence, there is an urgent need for effective multimodal rumor detection methods. However, existing approaches often overlook the insufficient diversity of multimodal samples in feature space and hidden similarities and differences among multimodal samples. To address such challenges, we propose MVACLNet, a Multimodal Virtual Augmentation Contrastive Learning Network. In MVACLNet, we first design a Hierarchical Textual Feature Extraction (HTFE) module to extract comprehensive textual features from multiple perspectives. Then, we fuse the textual and visual features using a modified cross-attention mechanism, which operates from different perspectives at the feature value level, to obtain authentic multimodal feature representations. Following this, we devise a Virtual Augmentation Contrastive Learning (VACL) module as an auxiliary training module. It leverages ground-truth labels and extra-generated virtual multimodal feature representations to enhance contrastive learning, thus helping capture more crucial similarities and differences among multimodal samples. Meanwhile, it performs a Kullback–Leibler (KL) divergence constraint between predicted probability distributions of the virtual multimodal feature representations and their corresponding virtual labels to help extract more content-invariant multimodal features. Finally, the authentic multimodal feature representations are input into a rumor classifier for detection. Experiments on two real-world datasets demonstrate the effectiveness and superiority of MVACLNet on multimodal rumor detection.

Topic to Image: A Rumor Detection Method Inspired by Image Forgery Recognition Technology

Topic to Image: A Rumor Detection Method Inspired by Image Forgery Recognition Technology

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

Abstract In the rumor detection based on crowd intelligence, the crowd behavior is constructed as a graph model or probability mode. The detection of rumors is achieved through the collaborative utilization of data and knowledge. Aiming at the problems of insufficient feature extraction ability and data redundancy of current rumor detection methods based on deep learning model, a social network rumor detection method based on bidirectional gated recurrent unit (Bi-GRU) and double self-attention (DSA) mechanism is suggested. First, a combination of application program interface and third-party crawler approach is used to obtain microblogging data from publicly available fake microblogging information pages, including both rumor and non-rumor information. Second, Bi-GRU is used to capture the tendency of medium- and long-term dependence of data and is flexible enough to deal with variable length input. Finally, the DSA mechanism is introduced to help reduce the redundant information in the dataset, thereby enhancing the model’s efficacy. The results of the experiments indicate that the proposed method outperforms existing advanced methods by at least 0.114, 0.108, 0.064, and 0.085 in terms of accuracy, precision, recall, and F1-scores, respectively. Therefore, the proposed method can significantly enhance the ability of social network rumor detection.

Harmfulness metrics in digital twins of social network rumors detection in cloud computing environment

Social network rumor harm metric is a task to score the harm caused by a rumor by analyzing the spreading range of the rumor, the users affected, the repercussions caused, etc., and then the harm caused by the rumor. Rumor hazard metric models can help rumor detection digital twins to understand and analyze user behaviors and assist social network network managers to make more informed decisions. However, there is a lack of models that can quantify the harm of rumors and automated harm metric models in rumor detection digital twins. To address this issue, this paper proposes an innovative social network rumor harm metric based on rumor propagation knowledge and a large language model (LLM), RSK-T5. The method first completes the joint task of rumor comment stance detection and sentiment analysis to capture critical features of rumor propagation. Then, this knowledge is used in the pre-training process of LLM to improve the model's understanding of rumor propagation patterns. Finally, the fine-tuning phase focuses on the hazard metrics task to improve the generalization energy. We compare with some existing variants of rumor detection methods, and experimental results demonstrate that RSK-T5 achieves the lowest MSE scores on three well-known rumor detection datasets. The ablative learning work demonstrates the effectiveness of RSK-T5's knowledge of two rumor spreads.

SAMGAT: structure-aware multilevel graph attention networks for automatic rumor detection.

The rapid dissemination of unverified information through social platforms like Twitter poses considerable dangers to societal stability. Identifying real versus fake claims is challenging, and previous work on rumor detection methods often fails to effectively capture propagation structure features. These methods also often overlook the presence of comments irrelevant to the discussion topic of the source post. To address this, we introduce a novel approach: the Structure-Aware Multilevel Graph Attention Network (SAMGAT) for rumor classification. SAMGAT employs a dynamic attention mechanism that blends GATv2 and dot-product attention to capture the contextual relationships between posts, allowing for varying attention scores based on the stance of the central node. The model incorporates a structure-aware attention mechanism that learns attention weights that can indicate the existence of edges, effectively reflecting the propagation structure of rumors. Moreover, SAMGAT incorporates a top-k attention filtering mechanism to select the most relevant neighboring nodes, enhancing its ability to focus on the key structural features of rumor propagation. Furthermore, SAMGAT includes a claim-guided attention pooling mechanism with a thresholding step to focus on the most informative posts when constructing the event representation. Experimental results on benchmark datasets demonstrate that SAMGAT outperforms state-of-the-art methods in identifying rumors and improves the effectiveness of early rumor detection.

Rumor detection technology based on ubiquitous relationship

This paper addresses the limitations of existing rumor detection methods that heavily rely on single or local features, which restrict their ability to capture comprehensive and detailed characteristics of rumors. The main objective of this study is to enhance the efficiency of rumor detection. To achieve this, we propose a novel approach that integrates user attributes, comment structure, and propagation models, introducing the concept of ubiquitous relationships for messages in social networks. We construct a Tweet-word-user ubiquitous relationship network using a propagation model and further leverage the Graph Convolutional Neural Network (GCN) to enhance semantic features. Consequently, we present a novel rumor detection model, the Ubiquitous Relationship-based Graph Convolutional Neural Network (U-GCN), which effectively combines user, text, and comment features within a unified framework, while also enhancing semantic features from the source post for comprehensive detection. Extensive experiments are conducted on two publicly available Twitter Datasets. The results demonstrate that our proposed U-GCN model achieves an accuracy rate of above 0.9, outperforming methods that solely consider single or local features. Our findings highlight the effectiveness of leveraging ubiquitous relationships in rumor detection.

Multi-domain rumor detection method based on sentiment features and attention mechanism

The wide dissemination of rumor is increasingly threatening both individuals and society. In this paper, we propose a multi-domain rumor detection method based on sentiment features and attention mechanism. This method uses sentiment analysis technology to extract sentiment features from text, and uses attention mechanism to weight text features and domain features to detect rumor. Experimental results show that the proposed method has achieved good results on data sets from multiple fields and has good generalization performance.

BiMGCL: rumor detection via bi-directional multi-level graph contrastive learning

The rapid development of large language models has significantly reduced the cost of producing rumors, which brings a tremendous challenge to the authenticity of content on social media. Therefore, it has become crucially important to identify and detect rumors. Existing deep learning methods usually require a large amount of labeled data, which leads to poor robustness in dealing with different types of rumor events. In addition, they neglect to fully utilize the structural information of rumors, resulting in a need to improve their identification and detection performance. In this article, we propose a new rumor detection framework based on bi-directional multi-level graph contrastive learning, BiMGCL, which models each rumor propagation structure as bi-directional graphs and performs self-supervised contrastive learning based on node-level and graph-level instances. In particular, BiMGCL models the structure of each rumor event with fine-grained bidirectional graphs that effectively consider the bi-directional structural characteristics of rumor propagation and dispersion. Moreover, BiMGCL designs three types of interpretable bi-directional graph data augmentation strategies and adopts both node-level and graph-level contrastive learning to capture the propagation characteristics of rumor events. Experimental results on real datasets demonstrate that our proposed BiMGCL achieves superior detection performance compared against the state-of-the-art rumor detection methods.

EPRD: Exploiting prior knowledge for evidence-providing automatic rumor detection

With the prevalence of social media platforms, rumors have been a serious social problem. Notably, existing rumor detection methods simply provide detection labels while ignoring their explanation. However, illustrating the reasons why a suspicious statement is a rumor is essential. To address this realistic scenario, we propose a novel Evidence-Providing Rumor Detection model called EPRD. EPRD incorporates a wide variety of information from both prior knowledge sources and current comments. It also learns bilaterally friendly representations for interpretable rumor detection. EPRD first retrieves evidence from prior knowledge sources and checks the relationship between the given statement and its evidence. Our model then constructs two heterogeneous graph objects to simulate the propagation layout of the current comments and evidence relationships. Finally, EPRD integrates the GraphSAGE component and attention mechanism to detect rumors. To the best of our knowledge, we propose the first model that incorporates prior knowledge to verify rumors and boost credibility. Experiments on two real-world Twitter datasets demonstrate that EPRD consistently exhibits the best rumor detection performance. Moreover, EPRD outperforms other baselines in the early rumor detection task.

Dual-channel early rumor detection based on factual evidence

Social media facilitates people's free expression and communication, but it also provides a platform for the generation and dissemination of rumors. In order to curb the spread of rumors in time and minimize their harm, early rumor detection methods were born and became a research hotspot. Early rumor detection requires a balance between accuracy and timeliness, but most existing methods only focused on accuracy and neglected efficiency. In this regard, we analyze a large number of Internet rumors and found that some rumors have been repeatedly spread after a small amount of modification. Such old rumors can be quickly and easily identified through factual evidence, while other new rumors require more clues and time to crack. Based on this insight, we propose a dual-channel early rumor detection model named ERD-DC (Early Rumor Detection-Dual-Channel), which leverages deep learning algorithms. Specifically, ERD-DC consists of two distinct channels, a fast channel referred to as channel-E (channel-Easy) and a slow channel named channel-D (channel-Difficult). In the ERD-DC, posts are divided into ‘easy-to-identify’ and ‘difficult-to-identify’ based on the availability of relevant evidence online. Subsequently, both types of posts are directed to their respective channels for detection. Channel-E focuses on identifying easy-to-identify rumors through a straightforward and swift post-evidence matching process, primarily aimed at enhancing timeliness. In parallel, channel-D is tasked with uncovering difficult-to-identify rumors, leveraging a more comprehensive analysis of user behaviors and posts transmission structures to elevate overall accuracy. Experimental results on two real datasets demonstrate that ERD-DC outperforms other state-of-the-art models significantly in early rumor detection with an accuracy rate of 84%.

Social Network Rumor Detection Method Combining Dual-Attention Mechanism With Graph Convolutional Network

Most rumor detection methods extract the features of rumor through two aspects of text semantics and propagation structure to achieve automatic rumor classification, while most of the existing methods do not realize that false and irrelevant interactions in the propagation structure will reduce the accuracy of rumor detection. In addition, most of the existing rumor detection methods failed to effectively extract key clues from the comments of social network users. In response to these phenomena, this article proposes a social network rumor detection method combining a dual attention mechanism and graph convolutional network (GCN) (dual-attention GCN, DA-GCN). First, build an event propagation graph; then, the GCN is used to extract the propagation structure information of each event-related microblog (tweet), and the attention mechanism is combined to suppress the false and irrelevant interactive relationships. Therefore, the anti-interference propagation structure features are extracted from the propagation graph. Second, to fully utilize the clues in users’ comments, this article makes use of the attention mechanism to fuse source microblog (tweet) with the comment–retweet information and extract interactive semantic features from it. Finally, the above two features are fused to generate a new event representation. Experimental results show that the proposed DA-GCN has an accuracy of 94.4%, 90.5%, and 90.2% on the Weibo dataset, the Twitter15 dataset, and the Twitter16 dataset, respectively, and has achieved excellent performance in the early rumor detection task, which proves that the proposed method is reasonable and effective.

Rumor Detection in Social Media Based on Multi-Hop Graphs and Differential Time Series

The widespread dissemination of rumors (fake information) on online social media has had a detrimental impact on public opinion and the social environment. This necessitates the urgent need for efficient rumor detection methods. In recent years, deep learning techniques, including graph neural networks (GNNs) and recurrent neural networks (RNNs), have been employed to capture the spatiotemporal features of rumors. However, existing research has largely overlooked the limitations of traditional GNNs based on message-passing frameworks when dealing with rumor propagation graphs. In fact, due to the issues of excessive smoothing and gradient vanishing, traditional GNNs struggle to capture the interactive information among high-order neighbors when handling deep graphs, such as those in rumor propagation scenarios. Furthermore, previous methods used for learning the temporal features of rumors, whether based on dynamic graphs or time series, have overlooked the importance of differential temporal information. To address the aforementioned issues, this paper proposes a rumor detection model based on multi-hop graphs and differential time series. Specifically, this model consists of two components: the structural feature extraction module and the temporal feature extraction module. The former utilizes a multi-hop graph and the enhanced message passing framework to learn the high-order structural features of rumor propagation graphs. The latter explicitly models the differential time series to learn the temporal features of rumors. Extensive experiments conducted on multiple real-world datasets demonstrate that our proposed model outperforms the previous state-of-the-art methods.

Graph Convolutional-Based Deep Residual Modeling for Rumor Detection on Social Media

The popularity and development of social media have made it more and more convenient to spread rumors, and it has become especially important to detect rumors in massive amounts of information. Most of the traditional rumor detection methods use the rumor content or propagation structure to mine rumor characteristics, ignoring the fusion characteristics of the content and structure and their interaction. Therefore, a novel rumor detection method based on heterogeneous convolutional networks is proposed. First, this paper constructs a heterogeneous map that combines both the rumor content and propagation structure to explore their interaction during rumor propagation and obtain a rumor representation. On this basis, this paper uses a deep residual graph convolutional neural network to construct the content and structure interaction information of the current network propagation model. Finally, this paper uses the Twitter15 and Twitter16 datasets to verify the proposed method. Experimental results show that the proposed method has higher detection accuracy compared to the traditional rumor detection method.

NIT: Searching for rumors in social network through neighborhood information transmission

Rumor detection has been a hot issue in current online public opinion governance. Most of the recent rumor detection methods tend to be supervised learning, while research about unsupervised detection methods is relatively rare, which focuses on deep learning models. For both interpretability and high performance, an unsupervised rumor detection method based on neighborhood information transmission (NIT) is proposed. Firstly, the initial importance regarded as the initial information of the object is evaluated from the perspective of local and global heterogeneous distance. Then the neighborhood information graph is constructed based on the neighborhood relationship on mixed data. On the neighborhood graph, NIT performs information transmission between neighbors and constantly updates the importance score by paying attention to the correlation between objects. Finally, the rumor object can be sought out by importance score ranking. In addition, to enhance the rumor identification effect, a propagation-based enhancement strategy is proposed to improve the performance of the unsupervised rumor detection algorithm by the propagation index of each object. Experiments were conducted on two UCI datasets and Weibo real-world rumor dataset, and the results proved the effectiveness of the NIT algorithm and the feasibility of the enhancement strategy.

Rumor detection on social networks focusing on endogenous psychological motivation

Emerging information acquisition methods represented by social networks are increasingly popular. The freedom and concealment make social media become the main platform for rumor spreading. Rumor detection has attracted the attention of both academia and industry, and a series of rumor detection methods have emerged. However, existing methods focus on modeling exogenous features such as text content and user characteristics, while ignoring the endogenous psychological motivation of users. Sociological and psychological researches on rumor have demonstrated the correlation between the endogenous psychology of users and their behavior in social networks. Therefore, this study focuses on the intrinsic psychological motivation of users, probes the psychological changes of users after they are exposed to rumors from the perspectives of active sharing and passive response, analyzes and judges user behaviors to explore an efficient rumor detection method. Among them, active sharing behavior is based on the investigation of social users’ motivation, decision making and goal setting, and is related to users’ motivation to actively participate in content by creating posts and so on. Passive response, which is closely related to personal emotions originating from the influence of external factors, is a study of users’ cognitive processes and is associated with the way they handle information without active participation. Experiments demonstrate the superiority of our proposed method in rumor detection tasks, improving accuracy by 2.1% over the current baseline on the Twitter16 dataset and enabling early detection of rumors as they emerge.

Unsupervised Cross-Domain Rumor Detection with Contrastive Learning and Cross-Attention

Massive rumors usually appear along with breaking news or trending topics, seriously hindering the truth. Existing rumor detection methods are mostly focused on the same domain, thus have poor performance in cross-domain scenarios due to domain shift. In this work, we propose an end-to-end instance-wise and prototype-wise contrastive learning model with cross-attention mechanism for cross-domain rumor detection. The model not only performs cross-domain feature alignment, but also enforces target samples to align with the corresponding prototypes of a given source domain. Since target labels in a target domain are unavailable, we use a clustering-based approach with carefully initialized centers by a batch of source domain samples to produce pseudo labels. Moreover, we use a cross-attention mechanism on a pair of source data and target data with the same labels to learn domain-invariant representations. Because the samples in a domain pair tend to express similar semantic patterns especially on the people’s attitudes (e.g., supporting or denying) towards the same category of rumors, the discrepancy between a pair of source domain and target domain will be decreased. We conduct experiments on four groups of cross-domain datasets and show that our proposed model achieves state-of-the-art performance.

Multimodal Fusion with Dual-Attention Based on Textual Double-Embedding Networks for Rumor Detection

Rumors may bring a negative impact on social life, and compared with pure textual rumors, online rumors with multiple modalities at the same time are more likely to mislead users and spread, so multimodal rumor detection cannot be ignored. Current detection methods for multimodal rumors do not focus on the fusion of text and picture-region object features, so we propose a multimodal fusion neural network TDEDA (dual-attention based on textual double embedding) applied to rumor detection, which performs a high-level information interaction at the text–image object level and captures visual features associated with keywords using an attention mechanism. In this way, we explored the ability to enhance feature representation with assistance from different modalities in rumor detection, as well as to capture the correlations of the dense interaction between images and text. We conducted comparative experiments on two multimodal rumor detection datasets. The experimental results showed that TDEDA could reasonably handle multimodal information and thus improve the accuracy of rumor detection compared with currently relevant multimodal rumor detection methods.

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

Currently, social media is full of rumors. To stop rumors from spreading further, rumor detection has received increasing attention. Recent rumor detection methods treat all propagation paths and all nodes on the paths as equally important, resulting in models that fail to extract the key features. In addition, most methods ignore user features, leading to limitations in the performance improvement of rumor detection. To address these problems, we propose a Dual-Attention Network model on propagation Tree structures named DAN-Tree, where a node-and-path dual-attention mechanism is designed to organically fuse deep structure and semantic information on the propagation structures of rumors, and path oversampling and structural embedding are employed to enhance the learning of deep structures. Finally, we deeply integrate user profiles into the propagation trees in DAN-Tree, thus proposing the DAN-Tree++ model to further improve performance. Empirical studies on four rumor datasets have shown that DAN-Tree outperforms the state-of-the-art rumor detection models learning on propagation structures, and the results on two datasets with user information validate the superior performance of DAN-Tree++ over other models using both user profiles and propagation structures. What's more, DAN-Tree, especially DAN-Tree++, has achieved the best performance on early detection tasks.

A review of rumor detection techniques in social networks

Social networks have accelerated the speed and scope of information dissemination. However, the lack of regulation and freedom of speech on social platforms has resulted in the widespread dissemination of the unverified message. Therefore, rapid and effective detection of social network rumors is essential to purify the network environment and maintain public security. Currently, the defects of rumor detection technology are that the detection time is too long and the timeliness is poor. In addition, the differences based on specific regions or specific fields will lead to deviations in the training dataset. In this paper, firstly, the definition of rumor is described, and the current problems and detection process of rumor detection are described; Secondly, introduce different data acquisition methods and analyze their advantages and disadvantages; Thirdly, according to the development of rumor detection technology, the existing rumor detection methods of artificial, machine learning and deep learning are analyzed and compared; Finally, the challenges of social network rumor detection technology are summarized.