A topological framework to explore longitudinal social networks
A topological framework to explore longitudinal social networks
复制标题
DOI：
10.1007/s10588-014-9176-3
复制
发表时间：
2014-09
影响因子：
1.8
通讯作者：
M. S. Uddin;Arif Khan;L. Hossain;Piraveenan Mahendra;S. Carlsson
中科院分区：
文献类型：
--
作者：
M. S. Uddin;Arif Khan;L. Hossain;Piraveenan Mahendra;S. Carlsson
纵向网络通过在一组参与者（例如，个人或组织）。纵向网络可以被视为单个静态网络（即，网络的结构是固定的），其聚集在某个时间段上观察到的所有边或者作为在整个网络观察时段上在不同时间点观察到的一系列静态网络（即，网络结构随着时间的推移而变化）。对纵向网络潜在结构变化的理解以及个体行为者对这些变化的贡献使研究人员能够研究此类网络的不同结构特性。通过遵循拓扑方法（即，静态拓扑和动态拓扑），本文首先提出了一个纵向社交网络的分析框架。在静态拓扑结构下，对整个观测期的聚合网络进行了社交网络分析（SNA）。较小的网络数据段（即，短间隔网络）在与整个网络观察周期相比更短的时间内累积，用于动态拓扑中的分析目的。在此框架下，本研究对两个纵向网络进行拓扑分析，以探索这两个网络在不同阶段随时间的行为者层面动态。所提出的拓扑框架可以用于探索各种纵向社交网络（例如，疾病传播网络和计算机病毒网络）。这将最终导致对此类网络的更好授权和控制。对于网络科学研究人员来说，这个框架将带来新的研究机会，以增强我们目前对不同方面的知识（例如，网络解体和个人行动者对网络演化的贡献）的纵向社交网络。
Longitudinal networks evolve over time through the creation and/or deletion of links among a set of actors (e.g., individuals or organizations). A longitudinal network can be viewed as a single static network (i.e., structure of network is fixed) that aggregates all the edges observed over some time period or as a series of static networks observed in different point of time over the entire network observation period (i.e., structure of network is changing over time). The understanding of the underlying structural changes of longitudinal networks and contributions of individual actors to these changes enable researchers to investigate different structural properties of such networks. By following a topological approach (i.e.,static topologyanddynamic topology), this paper first proposes a framework to analyze longitudinal social networks. Instatic topology, social networks analysis (SNA) methods are applied to theaggregatednetwork of entire observation period. Smaller segments of network data (i.e.,short-intervalnetwork) that are accumulated in less time compared to the entire network observation period are used in thedynamic topologyfor analysis purposes. Based on this framework, this study then conducts topological analysis of two longitudinal networks to explore over time actor-level dynamics during different phases of these two networks. The proposed topological framework can be utilized to explore structural vulnerabilities and evolutionary trend of various longitudinal social networks (e.g., disease spread network and computer virus network). This will eventually lead to better authorization and control over such networks. For network science researchers, this framework will bring new research opportunities to enhance our present knowledge about different aspects (e.g., network disintegration and contribution of individual actor’s to network evolution) of longitudinal social networks.