Visual Analytics of Time Dependent 2D Point Clouds

Two dimensional point data can be considered one of the most basic, yet one of the most ubiquitous data types
arising in a wide variety of applications. The basic scatter plot approach is widely used and highly effective for
data sets of small to moderate size. However, it shows scalability problems for data sets of increasing size, of
multiple classes and of time-dependency. In this short paper, we therefore present an improved visual analysis of
such point clouds. The basic idea is to monitor certain statistical properties of the data for each point and for each
class as a function of time. The output of the statistic analysis is used for identification of interesting data views
decreasing information overload. The data is interactively visualized using various techniques. In this paper, we
specify the problem, detail our approach, and present application results based on a real world data set.