Structuring Cognitive Information for Software Complexity Measurement

Cognitive complexity of software has emerged as an interesting research area recently with the growing studies in cognitive informatics disciplines. Cognitive complexity measurement attempts to quantify the software from the perspective of the difficulty for human brain to process and comprehend the software, in order to enable more precise prediction of the critical information about testability, reliability, and maintainability, as well as the effort spent in the software project. Cognitive informatics theories suggest that cognitive complexity of software depends on fundamental factors such as inputs, outputs, loops/branches structure, and number of operators and operands. Analysis in this paper shows the significant flaw of current cognitive complexity measures that they quantify the factors without considering the dependencies among them. We therefore propose a new method to solve this problem by structuring the factors. The proposed measure was evaluated comparatively to existing metrics, and also proven by satisfying all nine Weyuker´s properties.