Applied Artificial Optimization Algorithm in Design Flaws Detection

The detection of design flaws is one of the most important aspects of software quality control, and the process should therefore be an integral part of the development and also the maintenance of software. It is possible to lower costs and extend the useful like cycle of the software simply by detecting design flaws at an early stage, and therefore attempts have been made to automate the procedures involved in detecting and fixing these flaws. One of the most common ways of detecting flaws is through the use of heuristic metrics which use predetermined standards as a means of analyzing the findings. While the approach can work successfully, the problem lies in the determination of those standards, or thresholds. This research study seeks to develop an enhanced method to improve threshold determination to be applied in flaw detection using metric-based designs. Accordingly, for each metric, an algorithm was employed for optimization of the contribution metrics to determine the threshold. The model produced threshold values which could then be adjusted to fit the requirements of the software data input. The findings from the experiments revealed that this approach could generate more appropriate thresholds for application in this context. In addition, the technique was relatively simple and could be used with different software programming languages, reducing the implementation time, and eliminating the need for the specialist expert support which would have traditionally been required in metric-based detection methods.