Multiple applications deal with processing datasets of huge volume and complicated structure. The starting points of the data processing are usually files with pre-defined fields, created either manually or programatically. These unprocessed records, and the further data obtained through aggregation and transformation of the starting data create a multi-dimensional vector, from which the user can gain useful information only if we manage to make it transparent. The main idea behind this task is to map the values of these multi-dimensional vectors to visual properties which can be easily recognized, so the user seeing the generated image can recognize patterns between correlating data. The goal is to visualize a dataset with more than three dimensions as transparent as we can, using dimension reduction, spatial positioning and movements, three dimension mapping, colors, sizes and shapes.
The tasks of the student are the following:
• To meet and summarize the most commonly used visualization methods.
• To engineer and implement an application, which realizes the visualization methods present in the task description.
• To analyze the performance of the application, and evulate the usability of these techniques from this viewpoint.
• To process a real dataset, and visualize it in a form which can be used to obtain information by human users.