IEEE Trans Vis Comput Graph. 2020 Apr;26(4):1650-1660. doi: 10.1109/TVCG.2020.2969056. Epub 2020 Feb 5.
The formation of social groups is defined by the interactions among the group members. Studying this group formation process can be useful in understanding the status of members, decision-making behaviors, spread of knowledge and diseases, and much more. A defining characteristic of these groups is the pecking order or hierarchy the members form which help groups work towards their goals. One area of social science deals with understanding the formation and maintenance of these hierarchies, and in our work we provide social scientists with a visual analytics tool - PeckVis - to aid this process. While online social groups or social networks have been studied deeply and lead to a variety of analyses and visualization tools, the study of smaller groups in the field of social science lacks the support of suitable tools. Domain experts believe that visualizing their data can save them time as well as reveal findings they may have failed to observe. We worked alongside domain experts to build an interactive visual analytics system to investigate social hierarchies. Our system can discover patterns and relationships between the members of a group as well as compare different groups. The results are presented to the user in the form of an interactive visual analytics dashboard. We demonstrate that domain experts were able to effectively use our tool to analyze animal behavior data.
社会群体的形成是由群体成员之间的相互作用定义的。研究这个群体形成的过程对于理解成员的地位、决策行为、知识和疾病的传播等方面都非常有用。这些群体的一个定义特征是成员形成的啄序或等级制度，这有助于群体实现其目标。社会科学的一个领域致力于理解这些层次结构的形成和维护，在我们的工作中，我们为社会科学家提供了一个可视化分析工具——PeckVis，以帮助他们完成这个过程。虽然在线社交群体或社交网络已经被深入研究，并导致了各种分析和可视化工具的出现，但社会科学领域中对较小群体的研究缺乏合适工具的支持。领域专家认为，可视化他们的数据可以节省他们的时间，并揭示他们可能没有观察到的发现。我们与领域专家合作，构建了一个交互式可视化分析系统来研究社会层次结构。我们的系统可以发现群体成员之间的模式和关系，并比较不同的群体。结果以交互式可视化分析仪表板的形式呈现给用户。我们证明了领域专家能够有效地使用我们的工具来分析动物行为数据。
