Scalable visualization of model differences

Abstract

If large models are compared their difference can contain a huge number of local changes. Conventional methods for displaying differences cannot reasonably handle such large differences. This paper proposes a solution to this problem. Our approach is based on the concept of polymetric views and extends it in two ways: firstly, we propose metrics for differences which quantify properties of differences and distinguish relevant from irrelevant changes. Moreover, we propose new graphical features of polymetric views. This combination provides a scalable presentation of differences which makes the changes of large models comprehensible.