In the era of the Internet and big data, online social media platforms have been developing rapidly, which accelerate rumors circulation. Rumor detection on social media is a worldwide challenging task due to rumor’s feature of high speed, fragmental information and extensive range. Most existing approaches identify rumors based on single-layered hybrid features like word features, sentiment features and user characteristics, or multimodal features like the combination of text features and image features. Some researchers adopted the hierarchical structure, but they neither used rumor propagation nor made full use of its retweet posts. In this paper, we propose a novel model for rumor detection based on Graph Neural Networks (GNN), named Hierarchically Aggregated Graph Neural Networks (HAGNN). This task focuses on capturing different granularities of high-level representations of text content and fusing the rumor propagation structure. It applies a Graph Convolutional Network (GCN) with a graph of rumor propagation to learn the text-granularity representations with the spreading of events. A GNN model with a document graph is employed to update aggregated features of both word and text granularity, it helps to form final representations of events to detect rumors. Experiments on two real-world datasets demonstrate the superiority of the proposed method over the baseline methods. Our model achieves the accuracy of 95.7% and 88.2% on the Weibo dataset Ma et al. 2017 and the CED dataset Song et al. IEEE Trans Knowl Data Eng 33(8):3035–3047, 2019respectively.

Due to the wide availability of the tools used to produce manipulated images, a large number of digital images have been tampered with in various media, such as newspapers and social networks, which makes the detection of tampered images particularly important. Therefore, an image manipulation detection algorithm leveraged by the Faster Region-based Convolutional Neural Network (Faster R-CNN) model combined with edge detection was proposed in this paper. In our algorithm, first, original tampered images and their detected edges were sent into symmetrical ResNet101 networks to extract tampering features. Then, these features were put into the Region of Interest (RoI) pooling layer. Instead of the RoI max pooling approach, the bilinear interpolation method was adopted to obtain the RoI region. After the RoI features of original input images and edge feature images were sent into bilinear pooling layer for feature fusion, tampering classification was performed in fully connection layer. Finally, Region Proposal Network (RPN) was used to locate forgery regions. Experimental results on three different image manipulation datasets show that our proposed algorithm can detect tampered images more effectively than other existing image manipulation detection algorithms.

Social media has become more popular these days due to widely used instant messaging. Nevertheless, rumor propagation on social media has become an increasingly important issue. The purpose of this study is to investigate the impact of various features in social media on rumor detection, propose a dual co-attention-based multi-feature fusion method for rumor detection, and explore the detection capability of the proposed method in early rumor detection tasks. The proposed BERT-based Dual Co-attention Neural Network (BDCoNN) method for rumor detection, which uses BERT for word embedding. It simultaneously integrates features from three sources: publishing user profiles, source tweets, and comments. In the BDCoNN method, user discrete features and identity descriptors in user profiles are extracted using a one-dimensional convolutional neural network (CNN) and TextCNN, respectively. The bidirectional gate recurrent unit network (BiGRU) with a hierarchical attention mechanism is used to learn the hidden layer representation of tweet sequence and comment sequence. A dual collaborative attention mechanism is used to explore the correlation among publishing user profiles, tweet content, and comments. Then the feature vector is fed into classifier to identify the implicit differences between rumor spreaders and non-rumor spreaders. In this study, we conducted several experiments on the Weibo and CED datasets collected from microblog. The results show that the proposed method achieves the state-of-the-art performance compared with baseline methods, which is 5.2% and 5% higher than the dEFEND. The F1 value is increased by 4.4% and 4%, respectively. In addition, this paper conducts research on early rumor detection tasks, which verifies the proposed method detects rumors more quickly and accurately than competitors.

Social media makes it easy for individuals to publish and consume news, but it also facilitates the spread of rumors. This paper proposes a novel deep recurrent neural model with a symmetrical network architecture for automatic rumor detection in social media such as Sina Weibo, which shows better performance than the existing methods. In the data preparing phase, we filter the posts according to the followers of the user. We then use sequential encoding for the posts and multiple embedding layers to get better feature representation, and multiple recurrent neural network layers to capture the dynamic temporal signals characteristic. The experimental results on the Sina Weibo dataset show that: 1. the sequential encoding performs better than the term frequency-inverse document frequency (TF-IDF) or the doc2vec encoding scheme; 2. the model is more accurate when trained on the posts from the users with more followers; and 3. the model achieves superior improvements over the existing works on the accuracy of detection, including the early detection.

Many rumors spread quickly and widely on social media, affecting social stability. The rumors of most current detection methods only use textual information or introduce external auxiliary information (such as user information and propagation information) to enhance the detection effect, and the inherent statistical features of the corpus have not been fully used and compared with the external auxiliary features; in addition, statistical features are more certain and can only be obtained from textual information. Therefore, we adopted a method based on the adaptive fusion of word frequency distribution features and textual features to detect rumors. Statistical features were extracted by encoding statistical information through a variational autoencoder. We extracted semantic features and sequence features as textual features through a parallel network comprising a convolutional neural network and a bidirectional long-term memory network. In addition, we also designed an adaptive valve to only fuse useful statistical features with textual features according to the credibility of textual features, which can solve the over-fitting problem caused by the excessive use of statistical features. The accuracy of the model in three public datasets (Twitter15, Twitter16, and Weibo) reached 87.5%, 88.6%, and 95.8%, respectively, and the F1 value reached 87.4%, 88.5%, and 95.8%, respectively, showing that the model can effectively improve the performance of rumor detection.