Skip to main content

Multi-dimensional Data Analysis

Task Performance for Multidimensional Data Analysis

Analysing large, multi-dimensional data sets and extracting meaningful insight for decision making is a difficult and often impossible task without the right visual tools. A wide range of multi-dimensional visualisation techniques exist including Parallel Coordinates, star plots, glyphs, and scatterplot matrices. While of these techniques are frequently used for interactive visualisation of multi-dimentional data, it is not well understood how they compare against each other in terms of task performance.

In this project, we investigate multi-dimensional data perception and task performance through dedicated user studies. In a first study, we specifically focus on the relative performance of Parallel Coordinates and Scatterplot Matrices for solving simple tasks in various dimensions. In addition to analysing task accuracy and completion time, we study the viewing behaviour of analysts to better understand how different visual cues in the visual representation are being used for solving the given tasks.

Investigators

Dr. Rudolf Netzel (Lead), Visualization Research Center, University of Stuttgart
Dr. Jenny Vuong, Software and Computational Systems, CSIRO Data61
Dr. Ulrich Engelke, Decision Sciences, CSIRO Data61
Dr. Sean O’Donoghue, Software and Computational Systems, CSIRO Data61
Prof. Daniel Weiskopf, Visualization Research Center, University of Stuttgart
Dr. Julian Heinrich, Bayer Crop Science

Lifetime

2016 – ongoing