Measuring perceived clutter in concept diagrams

Clutter in a diagram can be broadly defined as how visually complex the diagram is. It may be that different users perceive clutter in different ways, however. Moreover, it has been shown that, for certain types of diagrams and tasks, an increase in clutter negatively affects task performance, making quantifying clutter an important problem. In this paper we investigate the perceived clutter in concept diagrams, a visual language used for representing ontologies. Using perceptual theory and existing research on clutter for other diagrams, we propose five plausible measures for assigning clutter scores to concept diagrams. By performing an empirical study we evaluated each of these proposed measures against participants' rankings of diagrams. Whilst more than one of our measures showed strong correlation with perceived clutter, our results suggest that a measure based on the number of points where lines cross is the most appropriate way to quantify clutter for concept diagrams.