THE PRELIMINARY ORAL EXAMINATION FOR THE DEGREE OF Ph.D. IN COMPUTER SCIENCE FOR

                             Cody Dunne

Network data structures have been used extensively in recent years for modeling entities and their ties, across such diverse disciplines as Computer Science, Sociology, Bioinformatics, Urban Planning, and Archeology. Analyzing network datasets involves understanding the complex relationships between entities, as well as any attributes, statistics, or groupings associated with them. A widely used class of visualizations called node-link diagrams excel at showing the network topology, attributes, and groupings simultaneously.

However, many existing node-link diagrams are not easily readable or difficult to extract meaning from because of (1) the inherent complexity of the relationships and (2) the number of items designers try to render in limited screen space. Moreover, for every network there are many potential unintelligible or even misleading drawings. Automated layout algorithms have helped, but frequently generate ineffective drawings even with an expert analyst at the helm. Other researchers and I have shown that there can be vast improvements in network drawings, but no one has found a single way to produce excellent drawings for all networks.

Since there is no single way forward for all networks, I propose to investigate three strategies that appear valuable. First, I plan to develop a set of readability metrics to measure the effectiveness of node-link diagrams and localize areas needing improvement. I will detail optimization recommendations for specific user tasks, in addition to several semi-automatic and automatic optimization techniques that will aid layouts of small networks. Second, I suggest reducing clutter by finding common network motifs and simplifying them with compact but representative glyphs. Third, I propose augmenting the layout of large, dense networks with attribute- and topology-based groupings. The design of these techniques will ensure they are easily comprehensible, unsurprising, and visually reversible.

My dissertation will contribute an understanding of why some node-link diagrams are difficult to read, what measures of readability could help guide designers and users, and several promising strategies for improving node-link diagram readability which demonstrate that progress is possible. This work will open several avenues of research, both in readability improvement and user education. 

Examining Committee:

Dr. Ben Shneiderman                 -          Chair

Dr. Carl Kingsford                     -          Dept’s Representative

Dr. Amitabh Varshney                -          Committee Member

EVERYBODY IS INVITED TO ATTEND THE PRESENTATION