Text in Diagrams: Challenges to and Opportunities of Automatic Layout.
- Zugl.: Kiel, Univ., Diss. 2019
Visual programming languages based on node-link diagrams are supposedly easy to use and to understand. This is only true, however, if the diagram elements are properly placed — a tedious and time-consuming process if done manually. Automatic graph layout algorithms alleviate users from that burden. Since even visual languages usually cannot make do without text, it follows that layout algorithms need to properly support textual labels. That is what this work is all about. We start by examining how enough space can be reserved for textual labels to be properly placed without overlaps.We then look at how users place comments in diagrams to establish relations to diagram elements. Our aim is to infer those, in order to take them into account during layout. We finally look at the negative implications of too much text: large diagrams and too much information. Different labelmanagement strategies dynamically change the text of labels, thus changing their size and, optionally, the amount of text displayed. All of the techniques are evaluated according to aesthetic criteria, and most are also validated through user studies.