The development of crowd intelligence makes the structure of social network more complex and changeable. Research on social network should be more in-depth and focus on the changes of structure. Ego-network, which represents the relationship between specific individual and the related people, is a hot issue among the research of dynamic social network. The evolution of ego-network is highly dynamic and pluralistic, it is hard to capture its evolutionary pattern over time. To help users analyze the individual characteristics and hidden patterns in multivariate ego-network, we present EgoVis, an interactive visual analysis system for exploring and analyzing complex structural relationships in dynamic network. Based on the task requirements of network evolution analysis, we propose a task taxonomy which is suitable for ego-network research and analysis, design novel visual fonts, and analyze the evolution of dynamic ego-network relations from the three dimensions: overview, subgroup, and detail-ego. Finally, the validity and practicability of EgoVis are verified on DBLP citation network dataset.

中文翻译：

---

EgoVis：基于自我中心研究的社交网络可视化分析系统

群体智能的发展使得社交网络的结构更加复杂多变。对社交网络的研究应该更加深入，关注结构的变化。自我网络，代表特定个体与相关人之间的关系，是动态社会网络研究的热点问题。自我网络的演变是高度动态和多元化的，很难捕捉到它随着时间的推移而演变的模式。为了帮助用户分析多元自我网络中的个体特征和隐藏模式，我们提出了一种交互式可视化分析系统 EgoVis，用于探索和分析动态网络中的复杂结构关系。基于网络演进分析的任务需求，我们提出了一种适用于自我网络研究和分析的任务分类法，设计新颖的视觉字体，并从三个维度分析动态自我网络关系的演变：概览、子组和细节自我。最后，在 DBLP 引文网络数据集上验证了 EgoVis 的有效性和实用性。