Authors: Darius Coelho, Sungsoo Ha, Shenghui Cheng, Salman Mahmood, Jisung Kim, Klaus Mueller
Abstract: We describe a framework that aids users in the creation of data visualizations augmented with chart junk. Chart junk refers to graphical decorations that are irrelevant to the data, but are meant to make the data graphics more interesting, giving rise to visualizations with low data-to-ink ratios. While outcast at first, recent studies have shown that users tend to find visualizations with chart junk more memorable, specifically for trends and the topic of the chart. It was found that images that contain human recognizable objects are especially useful. So far, chart junk creation has been a quite elaborate task, requiring the skills of an artist. With our MemViz tool now general users can also create attractive chart junk augmented visualizations. It is an interactive framework that links web-scale image search with a set of image processing tools to design chart junk data visualizations with ease.