Visualization of zoomable 2D projections on the web

The objective of the work is the research and development of a web-based visualization system for the creation and testing of zoomable projection cards. The basic idea is to project a multidimensional data set onto two dimensions using projection methods to represent it on a 2D surface. Based on the Card, Mackinlay, and Shneiderman visualization pipeline, a data processing model has been developed. For data processing various distance metrics, dimension reduction methods, zooming approaches as well as presentation concepts are considered. The peculiarities and considerations of the respective technology are discussed. A zooming approach allows large amounts of data to be displayed on a limited area. In order to better visualize connections within the data, concepts of presentation are discussed. The data points are represented as glyph-based objects or using color maps, various shapes, and sizes. Best practices about colormaps are discussed. In order to display large amounts of data in real time, a separation of the generation and visualization process takes place. During generation, a tabular file and selected configuration execute computationally-intensive transformation processes to create map material. Similar to Google Maps, the generated map material is represented by a visualization. Management concepts for managing various map sets as well as their generation and presentation are presented. A user interface can be used to create and visualize map material. The user uploads a tabular file into the system and chooses between different configuration parameters. Subsequently, this information is used to generate map material. The maps and various interaction options are provided in the visualization interface. Using various application examples, the advantages of this visualization system are presented.