Identifying fake news on media has been an important issue. This is especially true considering the wide spread of rumors on popular social networks such as Twitter. Various kinds of techniques have been proposed for automatic rumor detection. In this work, we study the application of graph neural networks for rumor classification at a lower level, instead of applying existing neural network architectures to detect rumors. The responses to true rumors and false rumors display distinct characteristics. This suggests that it is essential to capture such interactions in an effective manner for a deep learning network to achieve better rumor detection performance. To this end we present a simplified aggregation graph neural network architecture. Experiments on publicly available Twitter datasets demonstrate that the proposed network has performance on a par with or even better than that of state-of-the-art graph convolutional networks, while significantly reducing the computational complexity.

Synthetic aperture radar (SAR) image ship detection has been attracted attention widely. As a state-of-the-art method, constant false alarm rate (CFAR) detection algorithm is often used in SAR image ship detection applications. However, the detection results based on traditional sliding window in classical CFAR are very sensitive to the sliding window size, and it is difficult to avoid that there are no target pixels in clutter background. In addition, CFAR pixel-by-pixel detection algorithm is inefficient, especially for the detection of target in large sea scene high-resolution SAR images, because the computational complexity of CFAR algorithm based on sliding window is very high. To solve the above problems, this paper proposes a fast ship detection algorithm based on superpixel CFAR in SAR image. This method applies superpixels to CFAR target detection. Based on the superpixel segmentation results, the pure background clutter pixels of SAR image are obtained for clutter model parameter estimation, which avoids the shortcomings of sliding window design in traditional CFAR method. At the same time, compared with the traditional CFAR with the pixel-by-pixel clutter estimation, all the pixel clutter estimation parameters in each superpixel are unified, which greatly speeds up the detection. Experimental results show that the proposed method is very robust for ship detection in SAR images of different resolutions and scenes.

The recent development of the mobile Internet and the rise of social media have significantly enriched the way people access information. Accurate modeling of the probability of information propagation between users is essential for studying information dissemination issues in social networks. As the dissemination of information is inseparable from the interactions between users, the probability of propagation can be characterized by such interactions. In general, there are differences in the dissemination modes of information that carry different topics in a real social network. Using these factors, we propose a method (TMIVM) to measure the mutual influence between users at the topic level. The method associates two vectorization parameters for each user—an influence vector and a susceptibility vector—where the dimensions of the vector represent different topic categories. The magnitude of the mutual influence between users on different topics can be obtained by the product of the corresponding elements of the vectors. Specifically, in this article, we fit a social network historical information cascade data through Survival Analysis to learn the parameters of the influence and susceptibility vectors. The experimental results on a synthetic data set and a real Microblog data set show that this method better measures the propagation probability and information cascade predictions compared to other methods.

Firstly, the geometric model of linear sea surface is constructed based on linear wave theory and linear filtering method. At the same time, considering the nonlinear characteristics of sea waves, the modified model of nonlinear sea surface is introduced. By comparing the shape and statistical characteristics of nonlinear sea surface with linear sea surface, the Doppler echo spectrum of sea surface electromagnetic scattering is calculated, and the Doppler spectrum characteristics of dynamic sea surface scattering echo signals under different initial conditions are analyzed. Finally, the model simulation data are compared with the measured sea state data. The results show that the calculation results based on nonlinear sea surface are more consistent with the experimental data.

In this paper, we propose a new superpixel-based non-window CFAR ship detection method for high-resolution SAR images, which introduces the superpixels to CFAR detection. Firstly, our previously proposed fast Density Based Spatial Clustering of Applications with Noise (DBSCAN) superpixel generation method is utilized to produce the superpixels for high-resolution SAR images. With the assumption of SAR data obeying Gamma distribution, the superpixel dissimilarity is defined. Then superpixels can be used to accurately estimate the clutter parameters for the under test pixel, even in the multi-target situations, avoiding the drawbacks of sliding window in the traditional CFAR. Moreover, a local superpixel contrast is proposed to optimize the CFAR detection, which can eliminate the numerous man-made clutter false alarms, such as the urban areas. Experimental results with real SAR images indicate that the proposed method can achieve ship detection with a higher speed and accuracy in comparison with other state-of-the-art methods.

Identifying fake news on the media has been an important issue. This is especially true considering the wide spread of rumors on the popular social networks such as Twitter. Various kinds of techniques have been proposed for automatic rumor detection. In this work, we study the application of graph neural networks for rumor classification at a lower level, instead of applying existing neural network architectures to detect rumors. The responses to true rumors and false rumors display distinct characteristics. This suggests that it is essential to capture such interactions in an effective manner for a deep learning network to achieve better rumor detection performance. To this end we present a simplified aggregation graph neural network architecture. Experiments on publicly available Twitter datasets demonstrate that the proposed network has performance on a par with or even better than that of state-of-the-art graph convolutional networks, while significantly reducing the computational complexity.