Scalable Pixel based Visual Data Exploration

Abstract

Pixel based visualization techniques have proven to be of high value in visual data exploration. Mapping data points to pixels not only allows the analysis and visualization of large data sets, but also provides an intuitive way to convert raw data into a graphical form. The graphical representation fosters new insights and encourages the formation and validation of new hypotheses to the end of better problem solving and gaining deeper domain knowledge. However, the ever increasing amount of information leads to new challenges for pixel-based techniques and concepts, especially if the number of data points significantly exceeds the available screen resolution. The paper focuses on ways to improve the scalability of pixel based techniques by proposing a multiresolution pixel-oriented visual exploration approach for large datasets. This approach combines clustering techniques with pixel-oriented mappings to preserve local clusters while providing space filling relevancedriven representations of the whole data set or portions of the data. The paper presents different application scenarios from the fields of financial analysis, geo-visualization, and network data analysis that clearly show the practical benefit of the multi-resolution approach for tackling the problem of scalability.