Supporting the Visual Analysis of Dynamic Networks by Clustering associated Temporal Attributes
Supporting the Visual Analysis of Dynamic Networks by Clustering associated Temporal Attributes
复制标题
DOI：
10.1109/tvcg.2013.198
复制
发表时间：
2013-12
影响因子：
5.2
通讯作者：
S. Hadlak;H. Schumann;C. Cap;Till Wollenberg
中科院分区：
文献类型：
--
作者：
S. Hadlak;H. Schumann;C. Cap;Till Wollenberg
动态网络的可视化分析是一项具有挑战性的任务。在本文中，我们介绍了一种新的方法，通过结合计算，可视化和交互来支持发现具有相似趋势的子结构。使用现有的技术，由于需要比较的节点、边缘和时间点的数量，它们的发现将是一项乏味的工作。首先，在超图的基础上，我们根据随时间变化的相关属性对节点和边进行分组。其次，对超图进行可视化，以提供一组节点和边的概述，这些节点和边在一段时间内的行为与它们的相关属性相似。第三，我们提供了特定的交互来探索和完善时间聚类，允许用户进一步引导动态网络的分析。我们通过一个大型无线网状网络的可视化分析来演示我们的方法。
The visual analysis of dynamic networks is a challenging task. In this paper, we introduce a new approach supporting the discovery of substructures sharing a similar trend over time by combining computation, visualization and interaction. With existing techniques, their discovery would be a tedious endeavor because of the number of nodes, edges as well as time points to be compared. First, on the basis of the supergraph, we therefore group nodes and edges according to their associated attributes that are changing over time. Second, the supergraph is visualized to provide an overview of the groups of nodes and edges with similar behavior over time in terms of their associated attributes. Third, we provide specific interactions to explore and refine the temporal clustering, allowing the user to further steer the analysis of the dynamic network. We demonstrate our approach by the visual analysis of a large wireless mesh network.