Network Embedding Algorithm Research Articles

Tax Evasion Detection with FBNE-PU Algorithm Based on PnCGCN and PU Learning

Tax evasion is an illegal activity in which individuals or entities avoid paying their true tax liability. It has always been a crucial issue for both governments and academic researchers to efficiently detect tax evasion. Recent research has proposed the use of machine learning technology to detect tax evasion and has shown good results in some specific areas. Regrettably, there are still two major obstacles to detect tax evasion. First, it is hard to extract powerful features because of the complexity of tax data. Second, due to the complicated process of tax auditing, labeled data are limited. Such obstacles motivate the contributions of this work. In this paper, we propose a novel tax evasion detection framework named FBNE-PU, a multi-stage method to detect tax evasion in real-life scenarios. In this paper, we perform an in-depth analysis of the characteristics of the transaction network and propose a novel network embedding algorithm, the PnCGCN. It significantly improves detection performance by extracting powerful features from basic features and the tax-related transaction network. Moreover, we utilize nnPU to assign pseudo labels for unlabeled data. Finally, a MLP is trained as the decision function. Experiments on three real-world datasets demonstrate that our method significantly outperforms the comparison methods in the tax evasion detection task.

Conference Ranking based on Academic Network Embedding and Bipartite cross voting

Academic networks express the academic relationships among authors, papers and conferences, which offer perspectives and tools to rethink and solve traditional problems in the academic field. Among them, conference ranking is a problem which normally produced by collecting information from different sources and considering many factors. Not like most of the existed ranking methods that based on subjective evaluation, in this paper, we develop a Conference Ranking method based on Academic Network Embedding (CRANE) algorithm to simultaneously rank authors and conferences which improving the accuracy of ranking by embedding the Academic Network. The CRANE uses both conference features like internationality, citations and co-authors, and structure in Academic Network to evaluate the ranking of certain conferences. The theoretical analysis and convergence analysis were provided in this paper. Experiments on datasets of the Digital Bibliography & Library Project (DBLP) and AMiner academic platforms also be conducted. The experimental results show that the CRANE algorithm is superior to other algorithms by comparing with the public academic ranking like the China Computer Federation (CCF) ranking.

GCMD: Genetic Correlation Multi-Domain Virtual Network Embedding Algorithm

With the increase of network scale and the complexity of network structure, the problems of traditional Internet have emerged. At the same time, the appearance of network function virtualization (NFV) and network virtualization technologies has largely solved this problem, they can effectively split the network according to the application requirements, and flexibly provide network functions when needed. During the development of virtual network, how to improve network performance, including reducing the cost of embedding process and shortening the embedding time, has been widely concerned by the academia. Combining genetic algorithm with virtual network embedding problem, this paper proposes a genetic correlation multi-domain virtual network embedding algorithm (GCMD-VNE). The algorithm improves the natural selection stage and crossover stage of genetic algorithm, adds more accurate selection formula and crossover conditions, and improves the performance of the algorithm. Simulation results show that, compared with the existing algorithms, the algorithm has better performance in terms of embedding cost and embedding time.

Estimating the Political Orientation of Twitter Users Using Network Embedding Algorithms

Estimating the political orientation of citizens has always been a crucial task in communication as well as political science studies. In this study, advanced network analysis tools are developed to tackle this task. Specifically, using network embedding algorithms, friendship networks are embedded into lower-dimensional Euclidean space while preserving specific topological features of social networks. The resulting embedded vectors are then used to estimate political orientation of Twitter users. It is also shown that these numerical representations can be used to estimate other user traits. The developed tools are applied to a benchmark dataset as well as a dataset developed by the authors. Our model decreased the mean absolute error of the state-of-the-art predictions on the benchmark income dataset by 15%. The developed tools have multiple use cases, for example, studying echo chambers and political communication on OSNs and in marketing campaigns to estimate user’s preferences.

Distributed parallel genetic algorithm for online virtual network embedding

SummaryNetwork virtualization (NV) has emerged as a promising paradigm to address the constraints of implementing new protocols and services in existing network architecture by allowing the simultaneous coexistence of multiple heterogeneous virtual networks on a shared substrate infrastructure. Hence, NV is a critical technology for establishing future network architectures (e.g., 5G network and the smart Internet of Things [IoT]). Virtual network embedding (VNE) is a major challenge in NV since it is acknowledged as ‐hard. Many VNE solutions have been proposed over the past decade. However, the proposed solutions merely centralize VNE node mapping while recommending virtual link embedding for the shortest path method or multicommodity flow (MCF) mechanism. This research paper presents an intelligent virtual network orchestration based on genetic algorithm (GA) for the link mapping stage that implements distributed parallelism to significantly and efficiently reduce the operation time. Our extensive simulations have demonstrated that the proposed algorithm not only outperforms the state‐of‐the‐art VNE algorithm in all performance metrics but also achieves 44.01% faster embedding speed than the most well‐known, fastest link mapping method in VNE.

Sparse structure deep network embedding for transforming brain functional network in early mild cognitive impairment classification

AbstractCurrently, it remains one of the most challenging issues to distinguish brain functional networks of early mild cognitive impairment subjects (eMCIs) and normal control subjects (NCs). Unlike images, functional networks are non‐Euclidean data and not easily classified by dilated convolutional neural network (DCNN). To address this problem, we developed a sparse structure deep network embedding (SSDNE) method which transforms brain functional networks into a double‐channel image in eMCI classification. First, the eigenvector of each node in the functional network was obtained by SSDNE, and principal component analysis (PCA) was employed to sort the importance of eigenvectors. Next, two eigenvector groups with the highest contribution rate were extracted in turn and divided into several equal‐length intervals along its direction respectively, and the numbers of nodes that fall into the intervals were counted to obtain the corresponding two‐dimensional histograms. Then, the both histograms were stacked up and down into a double‐channel image. Finally, a double‐channel image was input into DCNN for feature learning to achieve final classification results. Experimental results show that, SSDNE performed better in maintaining the original structure of brain functional networks, and the transformed double‐channel image achieved comparably identifying results on eMCI classification compared with other network embedding algorithms. This novel method solved the problem that brain functional networks cannot be directly applied to convolutional neural networks for feature extraction and classification. Meanwhile, it can provide a reference for the early auxiliary diagnosis of Alzheimer's disease (AD).

Next waves in veridical network embedding*

AbstractEmbedding nodes of a large network into a metric (e.g., Euclidean) space has become an area of active research in statistical machine learning, which has found applications in natural and social sciences. Generally, a representation of a network object is learned in a Euclidean geometry and is then used for subsequent tasks regarding the nodes and/or edges of the network, such as community detection, node classification and link prediction. Network embedding algorithms have been proposed in multiple disciplines, often with domain‐specific notations and details. In addition, different measures and tools have been adopted to evaluate and compare the methods proposed under different settings, often dependent of the downstream tasks. As a result, it is challenging to study these algorithms in the literature systematically. Motivated by the recently proposed PCS framework for Veridical Data Science, we propose a framework for network embedding algorithms and discuss how the principles of predictability, computability, and stability (PCS) apply in this context. The utilization of this framework in network embedding holds the potential to motivate and point to new directions for future research.

Semisupervised Community Preserving Network Embedding with Pairwise Constraints

Network embedding aims to learn the low-dimensional representations of nodes in networks. It preserves the structure and internal attributes of the networks while representing nodes as low-dimensional dense real-valued vectors. These vectors are used as inputs of machine learning algorithms for network analysis tasks such as node clustering, classification, link prediction, and network visualization. The network embedding algorithms, which considered the community structure, impose a higher level of constraint on the similarity of nodes, and they make the learned node embedding results more discriminative. However, the existing network representation learning algorithms are mostly unsupervised models; the pairwise constraint information, which represents community membership, is not effectively utilized to obtain node embedding results that are more consistent with prior knowledge. This paper proposes a semisupervised modularized nonnegative matrix factorization model, SMNMF, while preserving the community structure for network embedding; the pairwise constraints (must-link and cannot-link) information are effectively fused with the adjacency matrix and node similarity matrix of the network so that the node representations learned by the model are more interpretable. Experimental results on eight real network datasets show that, comparing with the representative network embedding methods, the node representations learned after incorporating the pairwise constraints can obtain higher accuracy in node clustering task and the results of link prediction, and network visualization tasks indicate that the semisupervised model SMNMF is more discriminative than unsupervised ones.

Community detection in complex networks using network embedding and gravitational search algorithm

The structural and functional characteristic features of nodes can be analyzed by visualizing community structure in complex networks. Community detection helps us to detect nodes having similar behavior in a system and organize the network into a network of closely connected groups or modules. Network embedding technique represents the nodes of the input graph into vector space and preserves their inherent and topological features and can contribute significantly to various applications in network analysis. In this paper, we propose a novel community detection method using network embedding technique. Firstly, nodes of the graph are embedded in feature space of d dimensions, and then low-rank approximation is applied to avoid the results from being affected by noise or outliers. Further, k-means clustering is employed to find the centroids of the clusters in the network and followed by a gravitational search algorithm to improve the results of centroids of clusters. Finally, we compute the effectiveness of detected communities using different performance measures. Our method serves as a universal framework towards applying and bench-marking various embedding techniques in graphs for performing community detection. We perform the test using various evaluation criteria on several real-life and synthetic networks and the obtained result reveals the utility of the proposed algorithm.

IoV Scenario: Implementation of a Bandwidth Aware Algorithm in Wireless Network Communication Mode

Wireless network communication has developed rapidly in recent years, especially in the field of Internet of vehicles (IoV). However, due to the limitations of traditional network architecture, resource scheduling in wireless network environment is still facing great challenges. We focus on the urgent need of users for bandwidth resources in the IoV scenario under virtual network environment. This paper proposes a bandwidth aware multi domain virtual network embedding (BA-VNE) algorithm. The algorithm is mainly aimed at the problem that users need a lot of bandwidth in wireless communication mode, and solves the problem of bandwidth resource allocation from the perspective of virtual network embedding (VNE). In order to improve the performance of the algorithm, we introduce particle swarm optimization (PSO) algorithm to optimize the performance of the algorithm. In order to verify the effectiveness of the algorithm, we have carried out simulation experiments from link bandwidth, mapping cost and virtual network request (VNR) acceptance rate. The final results show that the proposed algorithm is better than other representative algorithms in the above indicators.

Domain Adaptive Network Embedding

Recent works reveal that network embedding techniques enable many machine learning models to handle diverse downstream tasks on graph-structured data. However, as previous methods usually focus on learning embedding for a single network, they cannot learn representations transferable on multiple networks. Hence, it is important to design a network embedding algorithm that supports downstream model transferring on different networks, known as domain adaptation. In this article, we propose a Domain Adaptive Network Embedding framework, which applies Graph Convolutional Network to learn transferable embedding. In DANE, nodes from multiple networks are encoded to vectors via a shared and aligned embedding space. The distribution of embedding on different networks are further aligned by Adversarial Learning Regularization. To achieve better performance in scenarios where labels are provided, DANE adopts a cross-entropy error term of the GCN framework and class centroid aligning method. Moreover, DANE's advantages in learning transferable network embedding can be guaranteed theoretically. Extensive experiments reflect that the proposed framework outperforms other well-recognized network embedding baselines in cross-network domain adaptation tasks, and the semi-supervised components improve the performance significantly.

MGA: Momentum Gradient Attack on Network

The adversarial attack methods based on gradient information can adequately find the perturbations, that is, the combinations of rewired links, thereby reducing the effectiveness of the deep learning model-based graph embedding algorithms, but it is also easy to fall into a local optimum. Therefore, this article proposes a momentum gradient attack (MGA) against the graph convolutional network (GCN) model, which can achieve more aggressive attacks with fewer rewiring links. Compared with directly updating the original network using gradient information, integrating the momentum term into the iterative process can stabilize the updating direction, which makes the model jump out of poor local optimum and enhances the method with stronger transferability. Experiments on node classification and community detection methods based on three well-known network embedding algorithms show that MGA has a better attack effect and transferability.

Susceptible-infected-spreading-based network embedding in static and temporal networks

Link prediction can be used to extract missing information, identify spurious interactions as well as forecast network evolution. Network embedding is a methodology to assign coordinates to nodes in a low-dimensional vector space. By embedding nodes into vectors, the link prediction problem can be converted into a similarity comparison task. Nodes with similar embedding vectors are more likely to be connected. Classic network embedding algorithms are random-walk-based. They sample trajectory paths via random walks and generate node pairs from the trajectory paths. The node pair set is further used as the input for a Skip-Gram model, a representative language model that embeds nodes (which are regarded as words) into vectors. In the present study, we propose to replace random walk processes by a spreading process, namely the susceptible-infected (SI) model, to sample paths. Specifically, we propose two susceptible-infected-spreading-based algorithms, i.e., Susceptible-Infected Network Embedding (SINE) on static networks and Temporal Susceptible-Infected Network Embedding (TSINE) on temporal networks. The performance of our algorithms is evaluated by the missing link prediction task in comparison with state-of-the-art static and temporal network embedding algorithms. Results show that SINE and TSINE outperform the baselines across all six empirical datasets. We further find that the performance of SINE is mostly better than TSINE, suggesting that temporal information does not necessarily improve the embedding for missing link prediction. Moreover, we study the effect of the sampling size, quantified as the total length of the trajectory paths, on the performance of the embedding algorithms. The better performance of SINE and TSINE requires a smaller sampling size in comparison with the baseline algorithms. Hence, SI-spreading-based embedding tends to be more applicable to large-scale networks.

A Node Embedding-Based Influential Spreaders Identification Approach

Node embedding is a representation learning technique that maps network nodes into lower-dimensional vector space. Embedding nodes into vector space can benefit network analysis tasks, such as community detection, link prediction, and influential node identification, in both calculation and richer application scope. In this paper, we propose a two-step node embedding-based solution for the social influence maximization problem (IMP). The solution employs a revised network-embedding algorithm to map input nodes into vector space in the first step. In the second step, the solution clusters the vector space nodes into subgroups and chooses the subgroups’ centers to be the influential spreaders. The proposed approach is a simple but effective IMP solution because it takes both the social reinforcement and homophily characteristics of the social network into consideration in node embedding and seed spreaders selection operation separately. The information propagation simulation experiment of single-point contact susceptible-infected-recovered (SIR) and full-contact SIR models on six different types of real network data sets proved that the proposed social influence maximization (SIM) solution exhibits significant propagation capability.

A Dynamic and Collaborative Multi-Layer Virtual Network Embedding Algorithm in SDN Based on Reinforcement Learning

Most of existing virtual network embedding (VNE) algorithms only consider how to construct virtual networks more efficiently on a physical infrastructure, without considering the possibility that the constructed virtual networks may be further virtualized to multiple smaller ones. We define the former scenario as single-layer VNE and the later as multi-layer VNE. As the increasing popularity of deploying large datacenter networks and wide area networks with Software Defined Network (SDN) architectures, it becomes a new requirement and possibility to provide multi-layer encapsulated network services for large tenants who have hierarchical organizational structures or need fine-grained service isolation. However, existing VNE algorithm are not specifically designed for the above requirement and not flexible enough to deal with mapping virtual network requirements (VNRs) to a physical network and smaller VNRs to a mapped virtual network. In this paper, we aim to propose a unified and flexible multi-layer VNE algorithm combining with reinforcement learning to solve the embedding of multi-layer VNRs, which can better distinguish the differences between VNRs and physical networks. Simulation results show that our algorithm achieves good performance both in single-layer and multi-layer VNE scenarios.

DSCD: Delay Sensitive Cross-Domain Virtual Network Embedding Algorithm

In the coming 5G era, more and more applications are time-delay sensitive, and the importance of low delay in fields such as telemedicine, autonomous driving and military order transmission is self-evident. However, for different application scenarios, their delay requirements are different. How to meet different Quality of service (QoS) requirements simultaneously becomes a challenge. The emergence of network virtualization (NV) technology can overcome this problem. However, it is difficult for the existing embedding algorithms to meet the requirements of virtual networks on time delay performance. Therefore, in order to study the delay problem of virtual network, we proposed a Delay Sensitive Cross-Domain Virtual Network Embedding Algorithm (DSCD-VNE) in this paper. First of all, different from the traditional virtual network embedding (VNE) algorithm, this algorithm divides the mapping process into a “node-link-node” three-stage embedding algorithm. Second, node resource metrics and set constraints are introduced in the generation phase of candidate link matrix. Third, the algorithm takes candidate link matrix as the core, and uses Kruskal minimum spanning tree to choose the link with the shortest delay for embedding. Fourth, path splitting mechanism and $K$ -shortest path algorithm are introduced with the aim of increasing the utilization of the substrate links in link mapping stage. The simulation results show that the proposed DSCD-VNE algorithm greatly improves the performance of the algorithm compared with the traditional node-link two-stage algorithm, especially in the aspect of delay, the improvement rate is as high as 77%.

MIPDH: A Novel Computational Model for Predicting microRNA-mRNA Interactions by DeepWalk on a Heterogeneous Network.

Analysis of miRNA-target mRNA interaction (MTI) is of crucial significance in discovering new target candidates for miRNAs. However, the biological experiments for identifying MTIs have a high false positive rate and are high-priced, time-consuming, and arduous. It is an urgent task to develop effective computational approaches to enhance the investigation of miRNA-target mRNA relationships. In this study, a novel method called MIPDH is developed for miRNA–mRNA interaction prediction by using DeepWalk on a heterogeneous network. More specifically, MIPDH extracts two kinds of features, in which a biological behavior feature is learned using a network embedding algorithm on a constructed heterogeneous network derived from 17 kinds of associations among drug, disease, and 6 kinds of biomolecules, and the attribute feature is learned using the k-mer method on sequences of miRNAs and target mRNAs. Then, a random forest classifier is trained on the features combined with the biological behavior feature and attribute feature. When implementing a 5-fold cross-validation experiment, MIPDH achieved an average accuracy, sensitivity, specificity and AUC of 75.85, 74.37, 77.33%, and 0.8044, respectively. To further evaluate the performance of MIPDH, other classifiers and feature descriptors are conducted for comparisons. MIPDH can achieve a better performance. Additionally, case studies on hsa-miR-106b-5p, hsa-let-7d-5p, and hsa-let-7e-5p are also implemented. As a result, 14, 9, and 9 out of the top 15 targets that interacted with these miRNAs were verified using the experimental literature or other databases. All these prediction results indicate that MIPDH is an effective method for predicting miRNA-target mRNA interactions.

6G technology based advanced virtual multi-purpose embedding algorithm to solve far-reaching network effects

At present, network virtualization is an essential technology that aims to resolve shortcomings including network ossification of the current design of the Internet. The Internet’s tremendous success have adopted digital business growth and increased competition for bandwidth in communication. Ultra-high-definition videos and vehicle systems require rapid bandwidth rates and increase network connection capacity, respectively. The rapid advancement of network virtualization and 6G technologies is driven by high bandwidth and high speed parallel communication. This innovation often introduces new link allocation problems to existing substrate networks in a network virtualization environment. In order to solve this far-reaching network effects, this paper proposes an Advanced Virtual Multi-Purpose Network Embedding Algorithm (AVMPNEA) through the fuzzy C-mean clustering, which is a learning algorithm. The clustering integrally evaluates the topological node structure, the latency and the related bandwidth between the nodes. The Nodes that enhance the longevity are used for selected mapping. The experimental result shows that the proposed AVMPNEA is outperformed in comparison with other traditional methods with high throughput, low latency and reduced energy consumption. Hence, it improves the throughput and the long-term application support.

PrGeFNE: Predicting disease-related genes by fast network embedding.

Identifying disease-related genes is of importance for understanding of molecule mechanisms of diseases, as well as diagnosis and treatment of diseases. Many computational methods have been proposed to predict disease-related genes, but how to make full use of multi-source biological data to enhance the ability of disease-gene prediction is still challenging. In this paper, we proposed a novel method for predicting disease-related genes by using fast network embedding (PrGeFNE), which can integrate multiple types of associations related to diseases and genes. Specifically, we first constructed a heterogeneous network by using phenotype-disease, disease-gene, protein-protein and gene-GO associations; and low-dimensional representation of nodes is extracted from the network by using a fast network embedding algorithm. Then, a dual-layer heterogeneous network was reconstructed by using the low-dimensional representation, and a network propagation was applied to the dual-layer heterogeneous network to predict disease-related genes. Through cross-validation and newly added-association validation, we displayed the important roles of different types of association data in enhancing the ability of disease-gene prediction, and confirmed the excellent performance of PrGeFNE by comparing to state-of-the-art algorithms. Furthermore, we developed a web tool that can facilitate researchers to search for candidate genes of different diseases predicted by PrGeFNE, along with the enrichment analysis of GO and pathway on candidate gene set. This may be useful for investigation of diseases' molecular mechanisms as well as their experimental validations. The web tool is available at http://bioinformatics.csu.edu.cn/prgefne/.

Protein2vec: Predicting Protein-Protein Interactions Based on LSTM.

The semantic similarity of gene ontology (GO) terms is widely used to predict protein-protein interactions (PPIs). The traditional semantic similarity measures are based mainly on manually crafted features, which may ignore some important hidden information of the gene ontology. Moreover, those methods usually obtain the similarity between proteins from similarity between GO terms by some simple statistical rules, such as MAX and BMA (best-match average), oversimplifying the possible complex relationship between the proteins and the GO terms annotated with them. To overcome the two deficiencies, we propose a new method named protein2vec, which characterizes a protein with a vector based on the GO terms annotated to it and combines the information of both the GO and known PPIs. We firstly try to apply the network embedding algorithm on the GO network to generate feature vectors for each GO term. Then, Long Short-Time Memory (LSTM) encodes the feature vectors of the GO terms annotated with a protein into another vector (called protein vector). Finally, two protein vectors are forwarded into a feedforward neural network to predict the interaction between the two corresponding proteins. The experimental results show that protein2vec outperforms almost all commonly used traditional semantic similarity methods.