Topological Features Of Nodes Research Articles

Heterogeneous question answering community detection based on graph neural network

Topic-based communities have gradually become a considerable medium for netizens to disseminate and acquire knowledge. These communities consist of entities (actual objects, e.g., a real answer or an actual question) with different types (users, questions and answers) and are usually hidden and overlapping. Nowadays, prevalent community question answering (CQA) platforms have formed mature communities by manually marked topics and extensive accumulated user behavior. However, the ever-growing various entities and complex overlapping topic communities make it inefficient to manually label entity tags (e.g., Question labels supplement domain features; Potential user tags indicate the user's specialty.). Therefore, there is an urgent need for a mechanism that automatically finds hidden semantic communities from user social behavior and lays a foundation for community construction and intelligent recommendation of QA platforms. In this paper, we propose a Heterogeneous Community Detection Approach Based on Graph Neural Network, called HCDBG, to detect heterogeneous communities in CQA. Firstly, we define entity relationships based on user interaction behavior and employ a heterogeneous information network to uniformly represent all connections. Afterward, we exploit the heterogeneous graph neural network to fuse content and topological features of nodes for graph embedding. Finally, we convert the community detection issue in CQA into an entity clustering task in the heterogeneous information network and improve the k-means method to achieve heterogeneous community detection. Based on our knowledge of the existing literature, it is an innovative research direction that utilizes the heterogeneous graph neural network to facilitate QA community detection. Extensive experiments on authentic question-answering datasets illustrate that HCDBG outperforms baseline methods in heterogeneous community detection.

Identification of the significant pathways of Banxia Houpu decoction in the treatment of depression based on network pharmacology.

Banxia Houpu decoction (BXHPD) has been used to treat depression in clinical practice for centuries. However, the pharmacological mechanisms of BXHPD still remain unclear. Network Pharmacology (NP) approach was used to explore the potential molecular mechanisms of BXHPD in treating depression. Potential active compounds of BXHPD were obtained from the Traditional Chinese Medicine Systems Pharmacology Database and Analysis Platform Database. STRING database was used to build a interaction network between the active compounds and target genes associated with depression. The topological features of nodes were visualized and calculated. Significant pathways and biological functions were identified using Gene Ontology and Kyoto Encyclopedia of Genes and Genomes analyses. A total of 44 active compounds were obtained from BXHPD, and 121 potential target genes were considered to be therapeutically relevant. Pathway analysis indicated that MAPK signaling pathway, ErbB signaling pathway, HIF-1 signaling pathway and PI3K-Akt pathway were significant pathways in depression. They were mainly involved in promoting nerve growth and nutrition and alleviating neuroinflammatory conditions. The result provided some potential ways for modern medicine in the treatment of depression.

Enhanced Affinity for Spectral Clustering using Topological Node Features (TNFS)

Data clustering is an active topic of research as it has applications in various fields such as biology, management, statistics, pattern recognition, etc. Spectral Clustering (SC) has gained popularity in recent times due to its ability to handle complex data and ease of implementation. A crucial step in spectral clustering is the construction of the affinity matrix, which is based on a pairwise similarity measure. The varied characteristics of datasets affect the performance of a spectral clustering technique. In this paper, we have proposed an affinity measure based on Topological Node Features (TNFs) viz., Clustering Coefficient (CC) and Summation index (SI) to define the notion of density and local structure. It has been shown that these features improve the performance of SC in clustering the data. The experiments were conducted on synthetic datasets, UCI datasets, and the MNIST handwritten datasets. The results show that the proposed affinity metric outperforms several recent spectral clustering methods in terms of accuracy.

Cluster-based topological features of nodes in a multiplex network—from a network of networks perspective

In the gathered multiplex systems, nodes inherit a part of their original system’s topological features, as in the world economic system, national policies and resource distribution bring industry advantages and resource advantages to the domestic industry. Although they represent one of the important original topological features of nodes, the inherited topological features of nodes have not received sufficient attention and have hardly been analyzed by existing network models. In our research, we defined the inherited topological features of nodes as ‘cluster-based topological features. To accurately calculate the cluster-based topological features of nodes in multiplex networks, we first provide a network model to summarize the multiplex networks into a calculable network of networks (NoN). Based on our network model, we propose a series of algorithms for calculating industries’ cluster-based topological features. Our calculating process contains 2 steps: ‘abstracting’ the NoN into one-layer calculable network; ‘inheriting’ subnetworks’ topological features into the inner nodes. Our network model and calculation algorithms are applied in a series of theoretical and empirical multiplex networks. The results not only confirm the realizability of our model but also produce several interesting findings, the most important of which is that some unremarkable nodes in multiplex network may have a very high contributory value from NoN perspective.

Recovery time after localized perturbations in complex dynamical networks

Maintaining the synchronous motion of dynamical systems interacting on complex networks is often critical to their functionality. However, real-world networked dynamical systems operating synchronously are prone to random perturbations driving the system to arbitrary states within the corresponding basin of attraction, thereby leading to epochs of desynchronized dynamics with a priori unknown durations. Thus, it is highly relevant to have an estimate of the duration of such transient phases before the system returns to synchrony, following a random perturbation to the dynamical state of any particular node of the network. We address this issue here by proposing the framework of single-node recovery time (SNRT) which provides an estimate of the relative time scales underlying the transient dynamics of the nodes of a network during its restoration to synchrony. We utilize this in differentiating the particularly slow nodes of the network from the relatively fast nodes, thus identifying the critical nodes which when perturbed lead to significantly enlarged recovery time of the system before resuming synchronized operation. Further, we reveal explicit relationships between the SNRT values of a network, and its global relaxation time when starting all the nodes from random initial conditions. Earlier work on relaxation time generally focused on investigating its dependence on macroscopic topological properties of the respective network. However, we employ the proposed concept for deducing microscopic relationships between topological features of nodes and their respective SNRT values. The framework of SNRT is further extended to a measure of resilience of the different nodes of a networked dynamical system. We demonstrate the potential of SNRT in networks of Rössler oscillators on paradigmatic topologies and a model of the power grid of the United Kingdom with second-order Kuramoto-type nodal dynamics illustrating the conceivable practical applicability of the proposed concept.

Botnet detection using graph-based feature clustering

Detecting botnets in a network is crucial because bots impact numerous areas such as cyber security, finance, health care, law enforcement, and more. Botnets are becoming more sophisticated and dangerous day-by-day, and most of the existing rule based and flow based detection methods may not be capable of detecting bot activities in an efficient and effective manner. Hence, designing a robust and fast botnet detection method is of high significance. In this study, we propose a novel botnet detection methodology based on topological features of nodes within a graph: in degree, out degree, in degree weight, out degree weight, clustering coefficient, node betweenness, and eigenvector centrality. A self-organizing map clustering method is applied to establish clusters of nodes in the network based on these features. Our method is capable of isolating bots in clusters of small sizes while containing the majority of normal nodes in the same big cluster. Thus, bots can be detected by searching a limited number of nodes. A filtering procedure is also developed to further enhance the algorithm efficiency by removing inactive nodes from consideration. The methodology is verified using the CTU-13 datasets, and benchmarked against a classification-based detection method. The results show that our proposed method can efficiently detect the bots despite their varying behaviors.

Introducing a procedure for developing a novel centrality measure (Sociability Centrality) for social networks using TOPSIS method and genetic algorithm

Centrality is one of the most important fields of social network research. To date, some centrality measures based on topological features of nodes in social networks have been proposed in which the importance of nodes is investigated from a certain point of view. Such measures are one dimensional and thus not feasible for measuring sociological features of nodes. Given that the main basis of Social Network Analysis (SNA) is related to social issues and interactions, a novel procedure is hereby proposed for developing a new centrality measure, named Sociability Centrality, based on the TOPSIS method and Genetic Algorithm (GA). This new centrality is not only based on topological features of nodes, but also a representation of their psychological and sociological features that is calculable for large size networks (e.g. online social networks) and has high correlation with the nodes' social skill questionnaire scores. Finally, efficiency of the proposed procedure for developing sociability centrality was tested via implementation on the Abrar Dataset. Our results show that this centrality measure outperforms its existing counterparts in terms of representing the social skills of nodes in a social network.

Efficient subgraph matching using topological node feature constraints

This paper presents techniques designed to minimise the number of states which are explored during subgraph isomorphism detection. A set of advanced topological node features, calculated from n-neighbourhood graphs, is presented and shown to outperform existing features. Further, the pruning effectiveness of both the new and existing topological node features is significantly improved through the introduction of strengthening techniques. In addition to topological node features, these strengthening techniques can also be used to enhance application-specific node labels using a proposed novel extension to existing pruning algorithms. Through the combination of these techniques, the number of explored search states can be reduced to near-optimal levels.