A Scalable Visualization for Dynamic Data in Software System Hierarchies

Abstract

Software systems can grow large, consisting of thousands of hierarchically organized elements like directories, subdirectories, files, and functions. Moreover, those hierarchy elements can carry additional information worth investigating for a software developer. Getting an overview of both the hierarchy and the attached static or dynamic data can become a tedious task if it is not supported by a visually scalable visualization technique. In this paper, we use a hierarchy visualization based on the visual metaphor of indentation to generate an overview of the software system hierarchy and easily attach additional attributes. The extra information is aligned with the hierarchy elements and, hence, supports visual comparisons of the attachments on different levels of hierarchical granularity. Through interaction, we provide additional views on the data, e.g., by filtering, hierarchy transformations, or details-on-demand. We illustrate the usefulness of our hierarchy visualization technique by means of an application example exploring data from the open-source software project jEdit. We investigated the readability of the hierarchy visualization with a user experiment, comparing indentation to node-link diagrams for varying sizes of a hierarchy.