Search

Recently network embedding has gained increasing attention due to its advantages in facilitating network computation tasks such as link prediction, node classification and node clustering. The objective of network embedding is to represent network nodes in a low-dimensional vector space while retaining as much information as possible from the original network including structural, relational, and semantic information. However, asymmetric nature of directed networks poses many challenges as how to best preserve edge directions in the embedding process. Here, we propose a novel deep asymmetric attributed network embedding model based on convolutional graph neural network, called AAGCN. The main idea is to maximally preserve the asymmetric proximity and asymmetric similarity of directed attributed networks. AAGCN introduces two neighbourhood feature aggregation schemes to separately aggregate the features of a node with the features of its in- and out- neighbours. Then, it learns two embedding vectors for each node, one source embedding vector and one target embedding vector. The final representations are the results of concatenating source and target embedding vectors. We test the performance of AAGCN on three real-world networks for network reconstruction, link prediction, node classification and visualization tasks. The experimental results show the superiority of AAGCN against state-of-the-art embedding methods., Comment: 16 Pages, 7 figures, 4 tables. Submitted to AEEE TNNLS

Vital node identification is the problem of finding nodes of highest importance in complex networks. This problem has crucial applications in various contexts such as viral marketing or controlling the propagation of virus or rumours in real-world networks. Existing approaches for vital node identification mainly focus on capturing the importance of a node through a mathematical expression which directly relates structural properties of the node to its vitality. Although these heuristic approaches have achieved good performance in practice, they have weak adaptability, and their performance is limited to specific settings and certain dynamics. Inspired by the power of machine learning models for efficiently capturing different types of patterns and relations, we propose a machine learning-based, data driven approach for vital node identification. The main idea is to train the model with a small portion of the graph, say 0.5% of the nodes, and do the prediction on the rest of the nodes. The ground-truth vitality for the train data is computed by simulating the SIR diffusion method starting from the train nodes. We use collective feature engineering where each node in the network is represented by incorporating elements of its connectivity, degree and extended coreness. Several machine learning models are trained on the node representations, but the best results are achieved by a Support Vector Regression machine with RBF kernel. The empirical results confirms that the proposed model outperforms state-of-the-art models on a selection of datasets, while it also shows more adaptability to changes in the dynamics parameters., Comment: 20 Pages, 5 figures, 6 tables. Submitted to Expert Systems with Applications

Identifying the most influential nodes in information networks has been the focus of many research studies. This problem has crucial applications in various contexts, such as controlling the propagation of viruses or rumours in real-world networks. While existing approaches mostly rely on the structural properties of networks and generate static rankings, in this work we propose a novel method that is responsive to any change in the diffusion dynamics. The main idea is to approximate the influential ability (influentiality) of a node with the reachability of other nodes from that node in a set of random sub-graphs. To this end, several random sub-graphs are sampled from the original network and then a hyper-graph is created in which each sub-graph is represented with a hyper-edge. From a theoretical standpoint, one can argue that a factor of the degree of nodes in the hyper-graph approximates influentiality. From an empirical perspective, the proposed model not only achieves the highest correlation with the ground-truth ranking, but also the ranking generated by this method hits the highest level of uniqueness and uniformity. Theoretical and practical analysis of the running time of this method also confirms a competitive running time compared with state-of-the-art methods., Comment: 18 Pages, 4 Figures, 5 Tables

Accessing to required data on the internet is wide via search engines in the last two decades owing to the huge amount of available data and the high rate of new data is generating daily. Accordingly, search engines are encouraged to make the most valuable existing data on the web searchable. Knowing how to handle a large amount of data in each step of a search engines' procedure from crawling to indexing and ranking is just one of the challenges that a professional search engine should solve. Moreover, it should also have the best practices in handling users' traffics, state-of-the-art natural language processing tools, and should also address many other challenges on the edge of science and technology. As a result, evaluating these systems is too challenging due to the level of internal complexity they have, and is crucial for finding the improvement path of the existing system. Therefore, an evaluation procedure is a normal subsystem of a search engine that has the role of building its roadmap. Recently, several countries have developed national search engine programs to build an infrastructure to provide special services based on their needs on the available data of their language on the web. This research is conducted accordingly to enlighten the advancement path of two Iranian national search engines: Yooz and Parsijoo in comparison with two international ones, Google and Bing. Unlike related work, it is a semi-automatic method to evaluate the search engines at the first pace. Eventually, we obtained some interesting results which based on them the component-based improvement roadmap of national search engines could be illustrated concretely.

Influence maximization is a widely studied topic in network science, where the aim is to reach the maximum possible number of nodes, while only targeting a small initial set of individuals. It has critical applications in many fields, including viral marketing, information propagation, news dissemination, and vaccinations. However, the objective does not usually take into account whether the final set of influenced nodes is fair with respect to sensitive attributes, such as race or gender. Here we address fair influence maximization, aiming to reach minorities more equitably. We introduce Adversarial Graph Embeddings: we co-train an auto-encoder for graph embedding and a discriminator to discern sensitive attributes. This leads to embeddings which are similarly distributed across sensitive attributes. We then find a good initial set by clustering the embeddings. We believe we are the first to use embeddings for the task of fair influence maximization. While there are typically trade-offs between fairness and influence maximization objectives, our experiments on synthetic and real-world datasets show that our approach dramatically reduces disparity while remaining competitive with state-of-the-art influence maximization methods., Comment: In Proc. of the 29th International Joint Conference on Artificial Intelligence (IJCAI'20), 2020