DyNetVis: a system for visualization of dynamic networks
DyNetVis: a system for visualization of dynamic networks
复制标题
DOI：
10.1145/3019612.3019686
复制
发表时间：
2017-04
期刊：
影响因子：
--
通讯作者：
Claudio D. G. Linhares;B. Travençolo;J. G. Paiva;Luis E C Rocha
中科院分区：
文献类型：
--
作者：
Claudio D. G. Linhares;B. Travençolo;J. G. Paiva;Luis E C Rocha
网络的概念在复杂系统的研究中一直很重要。在网络中，链路将成对的节点连接起来，形成复杂的结构。研究表明，网络不仅包含结构，还可能随着时间的推移而演变。时间维度的加入增加了分析的复杂性，并要求开发用于真实网络可视化的创新方法。本文介绍了动态网络可视化软件系统DyNetVis。该系统提供了几种用于用户交互的工具，并提供了两种协调的视觉布局，名为结构布局和时间布局。结构化是指标准网络绘制技术，其中节点和链路的单个快照被放置在平面中，而时间布局允许同时可视化动态网络的多个时间快照。此外，我们还研究了两种时态布局可视化方法：(I)循环邻居，一种在时间上突出频繁连接的节点排序策略；(Ii)时态活动图()，一种关注节点活动的布局技术。我们通过两个可视化分析案例说明了系统提供的布局和交互功能的适用性，展示了它们在改善动态网络上可视化和探索性数据分析的整体用户体验方面的优势。
The concept of networks has been important in the study of complex systems. In networks, links connect pairs of nodes forming complex structures. Studies have shown that networks not only contain structure but may also evolve in time. The addition of the temporal dimension adds complexity on the analysis and requests the development of innovative methods for the visualization of real-life networks. In this paper we introduce the Dynamic Network Visualization System (DyNetVis), a software tool for visualization of dynamic networks. The system provides several tools for user interaction and offers two coordinated visual layouts, named structural and temporal. Structural refers to standard network drawing techniques, in which a single snapshot of nodes and links are placed in a plane, whereas the temporal layout allows for simultaneously visualization of several temporal snapshots of the dynamic network. In addition, we also investigate two approaches for temporal layout visualization: (i) Recurrent Neighbors, a node ordering strategy that highlights frequent connections in time, and (ii) Temporal Activity Map (TAM), a layout technique with focus on nodes activity. We illustrate the applicability of the layouts and interaction functionalities provided by the system in two visual analysis case studies, demonstrating their advantages to improve the overall user experience on visualization and exploratory data analysis on dynamic networks.