A comprehensive appraisal of perceptual visual complexity analysis methods in GUI design

Graphical or Visual User Interface (GUI) is recognized as one of the most important application components for safety critical and business oriented software systems. It is highly advantageous for GUI designers and application developers to analyze the visual complexity of a GUI and predict users' perception and judgment during the design phase. Although in recent years, various methods have been developed for visual complexity analysis, these have not been widely used due to applicability, practicality and validity issues. In this respect, we have conducted a comprehensive review of studies and methods in visual complexity analysis. After identifying and analyzing 85 research studies, we grouped the visual complexity analysis methods and accordingly a taxonomy is presented. Furthermore, conceptual comparison of the methods is given and gap analysis as well as possible future directions are provided. According to the our findings, major gaps for each visual complexity analysis method may be stated as follows: 1) In metric-model based methods, there is a lack of information about the suitability of the metric-model created for analysis, since the extent to which each metric contributes to visual complexity analysis is still not known exactly. 2) In heuristic- based methods, the extracted rule set is not yet extendable enough beyond the use for specific GUIs. 3) While the visual complexity analysis could be considered as a kind of computer vision task, there exist limited studies that does so. Therefore, generalizable solutions based on machine learning techniques seem to be a promising research direction to develop efficient approaches.