Designing pixel-oriented visualization techniques: theory andapplications

摘要

Visualization techniques are of increasing importance in exploringnand analyzing large amounts of multidimensional information. Onenimportant class of visualization techniques which is particularlyninteresting for visualizing very large multidimensional data sets is thenclass of pixel-oriented techniques. The basic idea of pixel-orientednvisualization techniques is to represent as many data objects asnpossible on the screen at the same time by mapping each data value to anpixel of the screen and arranging the pixels adequately. A number ofndifferent pixel-oriented visualization techniques have been proposed innrecent years and it has been shown that the techniques are useful fornvisual data exploration in a number of different application contexts.nIn this paper, we discuss a number of issues which are important inndeveloping pixel-oriented visualization techniques. The major goal ofnthis article is to provide a formal basis of pixel-orientednvisualization techniques and show that the design decisions inndeveloping them can be seen as solutions of well-defined optimizationnproblems. This is true for the mapping of the data values to colors, thenarrangement of pixels inside the subwindows, the shape of thensubwindows, and the ordering of the dimension subwindows. The paper alsondiscusses the design issues of special variants of pixel-orientedntechniques for visualizing large spatial data sets