GenealogyVis: A System for Visual Analysis of Multidimensional Genealogical Data.
Yuhua Liu,Sicheng Dai,Changbo Wang,Zhiguang Zhou,Huamin Qu
DOI: https://doi.org/10.1109/thms.2017.2693236
2017-01-01
IEEE Transactions on Human-Machine Systems
Abstract:The study of genealogy is an increasingly popular activity pursued by millions of people, ranging from hobbyists to professional researchers. Such genealogical datasets provide a great opportunity for social science analysts, historians, and the public to study awide variety of topics in demography, family and household, kinship, stratification, and health. Nevertheless, the large scale and characteristics of the data such as hierarchical, spatiotemporal, and multidimensional also pose special challenges for effective data analysis. In this paper, we introduce GenealogyVis, a visual analytic system to analyze family history and evolution by using the China Multigenerational Panel Dataset-Liaoning, which has more than 1.5 million observations and provides socioeconomic, demographic, and other information for more than 260 000 residents, and further enable users to explore the correlation between the development of families and the social context of environments, economics, policies, and so on. This system includes five main linked views: the Scatter-plot View to provide an overview of the data and further explore the correlation analysis, the Tree View to show the family structure and details for individuals, the Migration View to present the genealogical migratory behaviors, the Matrix View to analyze the reproduction pattern between two generations, and the Stream View to show various statistical information such as demographic information and temporal information. A design study was conducted with a research group led by a domain expert of humanities and social sciences in an iterative manner over half a year. Several in-depth case studies, involving the research group, are described to demonstrate the usefulness of GenealogyVis and discuss new findings.