Bilkent University
Department of Computer Engineering


A hierarchical layout algorithm for compound graphs with non-uniform node dimensions


Asma Jodeiri Akbarfam
PhD Student
(Supervisor: Prof. Dr. Uğur Doğrusöz )
Computer Engineering Department
Bilkent University

Abstract: Over the years, the importance of graph visualization for understanding complex relationships among objects and the patterns in data has increased drastically. Effective visualization requires a decent layout of objects and their links for the user to effectively understand the data avoiding the cumbersome task of manual adjustments. Style and requirements for a good layout highly depend on the domain. However, most layout styles strive to achieve things like a minimal number of edge crossings, uniform edge lengths without sharp bends, and minimum area. These are usually in conflict with each other, making the problem extremely difficult. Domain-specific layouts favor some of these criteria over the others. Among many different layout styles, hierarchical layouts are popular for visual analysis of directed relationships in the form of hierarchical layers. Currently available hierarchical layout algorithms do a great job in satisfying the commonly accepted criteria for simple directed graphs with uniform node dimensions. However, none supports compound/nested graphs with non-uniform node dimensions. To address this, we would like to develop a new hierarchical layout algorithm that supports compound graphs and non-uniform node dimensions. The method will be based on the highly popular Sugiyama and Misue method.


DATE: 18 April 2022, Wednesday @ 16:20 Zoom