Selecting Coherent and Relevant Plots in Large Scatterplot Matrices

• Published in: Computer graphics forum Vol. 31; no. 6; pp. 1895 - 1908
• Main Authors: Lehmann, Dirk J., Albuquerque, Georgia, Eisemann, Martin, Magnor, Marcus, Theisel, Holger
• Format: Journal Article
• Language: English
• Published: Oxford, UK Blackwell Publishing Ltd 01.09.2012
• Subjects: Analysis; Computer graphics; H.3.3 [Information Search and Retrieval]: Information filtering; high-dimensional data; I.3.8 [Computer Graphics]: Applications; I.5.0 [Pattern Recognition]: Structural; quality measure; Scatter diagrams; scatterplot matrix; Studies; visual analytics
• ISSN: 0167-7055; 1467-8659
• DOI: 10.1111/j.1467-8659.2012.03069.x

The scatterplot matrix (SPLOM) is a well‐established technique to visually explore high‐dimensional data sets. It is characterized by the number of scatterplots (plots) of which it consists of. Unfortunately, this number quadratically grows with the number of the data set’s dimensions. Thus, an SPLOM scales very poorly. Consequently, the usefulness of SPLOMs is restricted to a small number of dimensions. For this, several approaches already exist to explore such ‘small’ SPLOMs. Those approaches address the scalability problem just indirectly and without solving it. Therefore, we introduce a new greedy approach to manage ‘large’ SPLOMs with more than 100 dimensions. We establish a combined visualization and interaction scheme that produces intuitively interpretable SPLOMs by combining known quality measures, a pre‐process reordering and a perception‐based ion. With this scheme, the user can interactively find large amounts of relevant plots in large SPLOMs. The scatterplot matrix (SPLOM) is a well‐established technique to visually explore high‐dimensional data sets. It is characterized by the number of scatterplots (plots) of which it consists of. Unfortunately, this number quadratically grows with the number of the data set's dimensions. Thus, an SPLOM scales very poorly. Consequently, the usefulness of SPLOMs is restricted to a small number of dimensions. For this, several approaches already exist to explore such ‘small’ SPLOMs. Those approaches address the scalability problem just indirectly and without solving it. Therefore, we introduce a new greedy approach to manage ‘large’ SPLOMs with more than 100 dimensions. We establish a combined visualization and interaction scheme that produces intuitively interpretable SPLOMs by combining known quality measures, a pre‐process reordering and a perception‐based ion.