The increase of cameras nowadays, and the power of the media in people's lives lead to a staggering amount of video data. It is certain that a method to process this large volume of videos quickly and optimally becomes especially important. With the help of video summarization, this task is achieved and the film is summarized into a series of short but meaningful frames or clips. This study tried to cluster the data by an algorithm (K-Medoids) and then with the help of a convolutional graph attention network, temporal and graph separation is done, then in the next step with the connection rejection method, noises and duplicates are removed, and finally summarization is done by merging the results obtained from two different graphical and temporal steps. The results were analyzed qualitatively and quantitatively on three datasets SumMe, TVSum, and OpenCv. In the qualitative method, an average of 88% accuracy rate in summarization and 31% error rate was achieved, which is one of the highest accuracy rates compared to other methods. In quantitative evaluation, the proposed method has a higher efficiency than the existing methods.

With the widespread use of Android smartphones, the Android platform has become an attractive target for cybersecurity attackers and malware authors. Meanwhile, the growing emergence of zero-day malware has long been a major concern for cybersecurity researchers. This is because malware that has not been seen before often exhibits new or unknown behaviors, and there is no documented defense against it. In recent years, deep learning has become the dominant machine learning technique for malware detection and could achieve outstanding achievements. Currently, most deep malware detection techniques are supervised in nature and require training on large datasets of benign and malicious samples. However, supervised techniques usually do not perform well against zero-day malware. Semi-supervised and unsupervised deep malware detection techniques have more potential to detect previously unseen malware. In this paper, we present MalGAE, a novel end-to-end deep malware detection technique that leverages one-class graph neural networks to detect Android malware in a semi-supervised manner. MalGAE represents each Android application with an attributed function call graph (AFCG) to benefit the ability of graphs to model complex relationships between data. It builds a deep one-class classifier by training a stacked graph autoencoder with graph convolutional layers on benign AFCGs. Experimental results show that MalGAE can achieve good detection performance in terms of different evaluation measures.

Background and Objectives: Twitter is a microblogging platform for expressing assessments, opinions, and sentiments on different topics and events. While there have been several studies around sentiment analysis of tweets and their popularity in the form of the number of retweets, predicting the sentiment of first-order replies remained a neglected challenge. Predicting the sentiment of tweet replies is helpful for both users and enterprises. In this study, we define a novel problem; given just a tweet's text, the goal is to predict the overall sentiment polarity of its upcoming replies.Methods: To address this problem, we proposed a graph convolutional neural network model that exploits the text's dependencies. The proposed model contains two parallel branches. The first branch extracts the contextual representation of the input tweets. The second branch extracts the structural and semantic information from tweets. Specifically, a Bi-LSTM network and a self-attention layer are used in the first layer for extracting syntactical relations, and an affective knowledge-enhanced dependency tree is used in the second branch for extracting semantic relations. Moreover, a graph convolutional network is used on the top of these branches to learn the joint feature representation. Finally, a retrieval-based attention mechanism is used on the output of the graph convolutional network for learning essential features from the final affective picture of tweets.Results: In the experiments, we only used the original tweets of the RETWEET dataset for training the models and ignored the replies of the tweets in the training process. The results on three versions of the RETWEET dataset showed that the proposed model outperforms the LSTM-based models and similar state-of-the-art graph convolutional network models. Conclusion: The proposed model showed promising results in confirming that by using only the content of a tweet, we can predict the overall sentiment of its replies. Moreover, the results showed that the proposed model achieves similar or comparable results with simpler deep models when trained on a public tweet dataset such as ACL 2014 dataset while outperforming both simple deep models and state-of-the-art graph convolutional deep models when trained on the RETWEET dataset. This shows the proposed model's effectiveness in extracting structural and semantic relations in the tweets.

In recent years, graph convolutional networks (GCNs) have achieved significant performance in the field of skeleton-based action recognition. Existing GCN-based methods usually apply fixed graph topologies and a temporal convolutional filter to extract the spatial and temporal features of an action. Since an action is coordinated through different parts of the body in the time domain and shows different characteristics in the time domain, this work causes the loss of a lot of information about an action.To address this issue, in this paper, we present an attention-based graph neural (AT-AR) network to discover distinct features from both spatial and temporal aspects. The proposed model uses an SPG Net convolution to learn Spatio-temporal features. In addition, the attention mechanism creates an attention score using temporal features, which can enhance the temporal correlations of an action. After establishing the two-stream structure, the AT-AR network has reached 96% and 97% accuracy under X-View and X-Sub on the NTU RGB+D dataset.

Today, the amount and importance of data available on the Internet is increasing exponentially, so choosing a suitable option from among many options can be tiring and time-consuming. The goal of recommendation systems is to facilitate this process by finding the right items that are of most interest to users. Existing recommendation systems suffer from common problems such as data sparsity, cold start, and new user problems. In this article, the main focus is on using information from other domains to create cross-domain recommendation systems. The proposed cross-domain systems can manage cold start situations and new users. In this article, first, a model based on convolutional graph neural networks discovers the interaction pattern of users and items in each domain independently, and in the next step, a neural network is used to transfer representation for cold-start users from the source domain to the target domain. The results show that the proposed model has a better performance compared to other proposed models for estimating the recorded score for the items.

This paper explores graph embedding techniques for effectively analyzing large, heterogeneous graphs with complex and noisy patterns. graphs represent data through nodes (entities) and edges (relationships), and when dealing with large-scale data, effective search methods are crucial. graph embedding helps evaluate node significance and transforms data into latent space representations. It also addresses challenges like handling multi-label data in heterogeneous networks, where nodes may have multiple labels describing complex concepts. Traditional methods struggle with such multi-label scenarios and fail to capture label dependencies. The paper introduces a graph neural Network (GCN)-based node embedding method, which extends traditional neural networks to graph data. GCNs allow the extraction of local features from nodes and their neighbors, making them useful for heterogeneous networks. By integrating label information into the embedding process, the method improves relationships between labels. The proposed approach transforms neighboring labels into continuous vectors, structured into a matrix for learning. This enhances the overall network embedding. The method outperforms previous techniques, demonstrating improved performance on real-world datasets, such as a 2.4% improvement on the IMDB dataset and 9.3% on the DBLP dataset. The paper discusses graph embedding techniques in the first section and explores the potential of multi-label embedding in non-uniform graphs, suggesting future research directions in the final section. The article's code link on GitHub can also be found at the following: https://github.com/frshkara/EGSA.

One of the standard criteria for expressing the relationship between two random variables is the correlation coefficient. Correlation between variables shows that changing the value of one variable leads to changing another variable in a certain direction. It is also possible to use the value of one variable to predict the value of another. In statistics, the correlation coefficient measures the direction and strength of the tendency to change. In machine learning, the correlation coefficient is known as a measure of classification quality. In fact, as a starting step for classification, the correlation between different samples should be estimated using a specific method. There are various methods to estimate the correlation of different data types, which have disadvantages such as low accuracy or high computational time. One of the methods that can overcome these problems, due to its high capability in modeling correlation between samples is graphical modeling. In this research, a new covariance model based on graph theory and graph neural network for estimating the correlation between samples is presented. The results show the improvement of the proposed model in accuracy, sensitivity, precision, F-Micro, F-Macro and statistical tests compared to Pearson and cosine methods.