Computing top-k temporal closeness in temporal networks

The closeness centrality of a vertex in a classical static graph is the reciprocal of the sum of the distances to all other vertices. However, networks are often dynamic and change over time. Temporal distances take these dynamics into account. In this work, we consider the harmonic temporal closeness with respect to the shortest duration distance. We introduce an efficient algorithm for computing the exact top-k temporal closeness values and the corresponding vertices. The algorithm can be generalized to the task of computing all closeness values. Furthermore, we derive heuristic modifications that perform well on real-world data sets and drastically reduce the running times. For the case that edge traversal takes an equal amount of time for all edges, we lift two approximation algorithms to the temporal domain. The algorithms approximate the transitive closure of a temporal graph (which is an essential ingredient for the top-k algorithm) and the temporal closeness for all vertices, respectively, with high probability. We experimentally evaluate all our new approaches on real-world data sets and show that they lead to drastically reduced running times while keeping high quality in many cases. Moreover, we demonstrate that the top-k temporal and static closeness vertex sets differ quite largely in the considered temporal networks.

The parallel computing of node centrality based on GPU

<abstract> <p>Many systems in real world can be represented as network, and network analysis can help us understand these systems. Node centrality is an important problem and has attracted a lot of attention in the field of network analysis. As the rapid development of information technology, the scale of network data is rapidly increasing. However, node centrality computation in large-scale networks is time consuming. Parallel computing is an alternative to speed up the computation of node centrality. GPU, which has been a core component of modern computer, can make a large number of core tasks work in parallel and has the ability of big data processing, and has been widely used to accelerate computing. Therefore, according to the parallel characteristic of GPU, we design the parallel algorithms to compute three widely used node centralities, i.e., closeness centrality, betweenness centrality and PageRank centrality. Firstly, we classify the three node centralities into two groups according to their definitions; secondly, we design the parallel algorithms by mapping the centrality computation of different nodes into different blocks or threads in GPU; thirdly, we analyze the correlations between different centralities in several networks, benefited from the designed parallel algorithms. Experimental results show that the parallel algorithms designed in this paper can speed up the computation of node centrality in large-scale networks, and the closeness centrality and the betweenness centrality are weakly correlated, although both of them are based on the shortest path.</p> </abstract>

The effect of network structure on the purchase of virtual goods on social networking services

PurposeThis paper aims to identify the effect of social structure variables on the purchase of virtual goods. Using field data, it also tests whether their effects on a social networking service are dynamic.Design/methodology/approachTo achieve the research objectives, the authors have applied the random effects panel Tobit model with actual time-series corporate data to explain a link between network structure factors and actual behavior on social networking services.FindingsThe authors have found that various network structure variables such as in-degree, in-closeness centrality, out-closeness centrality and clustering coefficients are significant predictors of virtual item sales; while the constraint is marginally significant, out-degree is not significant. Furthermore, these variables are time-varying, and the dynamic model performs better in a model fit than the static one.Practical implicationsThe findings will help social networking service (SNS) operators realize the importance of understanding network structure variables and personal motivations or the behavior of consumers.Originality/valueThis study provides implications in that it uses various and dynamic network structure variables with panel data.

Social Network Analysis: #BlackLivesMatter Distribution at Actor Level and System Level

#BlackLivesMatter accompanies several cases of discrimination against the black community. The hashtag was spread by actors who have great influences on Twitter users. The actors create communication network which connected to each other to form opinions about the Black Lives Matter movement. Researchers conducted a study to determine the distribution of #BlackLivesMatter at the actor level for the period 20-27 April 2021 in Twitter. The study used quantitative methods and a positivistic paradigm with a Social Network Analysis (SNA) approach. The results show that the actor with the highest degree of centrality is @jeanmessiha with 238 interactions, the actor with the highest betweenness centrality is @helloagain0611 with a value of 0.000049, the actor with the highest eigenvector centrality is @jeanmessiha with a value of 1 and there are 1,416 actors who have closeness centrality. # BlackLivesMatter has a low diameter value so that it spreads quickly but not too widely, not much reciprocity occurs, not concentrated in one dominant cluster but spread widely in several clusters. The actors play a role in spreading diverse opinions regarding Black Lives Matter, thus creating free discussion in several clusters on Twitter. Opinion widely spread on Twitter creates public opinion regarding the Black Lives Matter movement.

TPSL: A propagation source localization method based on limited observation nodes

The research on localization of propagation sources on complex networks has farreaching significance in various fields. Many source localization methods have been proposed. However, the assumptions of some existing methods are too ideal, which means they cannot be widely deployed on realistic networks. In this paper, we propose a multi-source localization method TPSL based on limited observation nodes and backward diffusion-based algorithm with the consideration of heterogeneity of the propagation probabilities between nodes. Specifically, given a network topology with time and probability distributions, TPSL can infer the sources of propagation by comprehensively considering the time and probability factors in a way that accords with the characteristics of information propagation in reality. The experiments on artificial and empirical networks demonstrate that TPSL has excellent performance on these networks. We also explore the influence of different strategies of choosing observation nodes on TPSL, and find out that choosing the nodes with larger closeness centrality as observation nodes performs better. Moreover, the performance of TPSL does not be affected by the number of sources.

The reception of support in peer-to-peer online networks: Network position, support solicitation, and support provision in an online asthma caregivers group

This study applies social network analysis and quantitative content analysis to messages exchanged within an online support forum of caregivers of children with chronic asthma to examine how peer-to-peer network positions and personal communication styles (seeking and providing support) impact the reception of social support. Content analysis is used to determine rates of giving and receiving informational and emotional support. Network analysis assesses levels of individual betweenness and closeness centrality in the online network. Relationships between network positions, solicitation strategies, and the provision and reception of informational and emotional support are examined. Betweenness and closeness centrality are associated with improved informational and emotional support. The provision of informational support is also improved by providing descriptions of personal experience. Practical implications for the design and use of online support platforms are discussed.

Social Network for Game of Thrones

Along with growing interest and use, the concept of network analysis has taken a new direction to explore data and facts to find existing patterns. The paper highlights the importance of social network analysis in analyzing and mining useful information from the data across various domains. It provides an insight into need, importance and scope of Social Network Analysis. With the use of Social networking tool like NetworkX, data is being represented in the form of graph or network which is then analyzed in a more efficient way making it easier to study the interactions between different persons in Game of thrones and establishing trends existing in a network. A comparative analysis of various centrality measures such as Degree centrality, Betweenness centrality, Closeness centrality, PageRank centrality is performed to explore the features associated to find the most important character of the series based on obtained results.

CDK1 and HSP90AA1 appears as novel regulatory gene in Non-Small Cell Lung Cancer: A Bioinformatics Approach

Lung cancer is one of the most invasive cancer affecting over a million of population. Non-small cell lung cancer constitutes up to 85% of all lung cancer cases. Therefore, it is important to identify prognostic biomarkers of NSCLC for therapeutic purpose. The complex behaviour of the NSCLC gene-regulatory network interaction is investigated using a network theoretical approach. We used eight NSCLC microarray datasets GSE19188, GSE118370, GSE10072, GSE101929, GSE7670, GSE33532, GSE31547, GSE31210 and meta analyse them to find differentially expressed genes (DEGs), construct protein-protein interaction (PPI) network, analysed its topological properties, significant modules using network analyser with MCODE, construct a PPI-MCODE network using the genes of the significant modules. We used topological properties such as Maximal Clique Centrality (MCC) and bottleneck from the PPI-MCODE network. We compare them with hub genes (those with highest degrees) to find key regulator (KR) gene. This result is also validated by finding of common genes among top twenty hub genes, genes with highest betweenness, closeness centrality and eigenvector values. It was found that two genes, CDK1 and HSP90AA1 were common in PPI-MCODE combined analysis, and it was also found that CDK1, HSP90AA1 and HSPA8 were common among hub and bottle neck properties and suggesting significant regulatory role of CDK1 in non-small cell lung cancer. After validation, the common genes among top twenty hubs and centrality values like Betweenness Centrality, Closeness Centrality and eigen vector properties, CDK1 again appeared as the common gene. Our study as a summary suggested CDK1 as key regulator gene in complex NSCLC network interaction using network theoretical approach and described the complex topological properties of the network.