The Collaborative Identification of Design Flaws in Software Systems

Abstract

Discovering defects in software source code is crucial to ensuring extendable and available computer systems, and a fundamental feature of their development and maintenance. Detecting such defects at an early stage of development can bring significant economic benefits, so efforts have been made to automate their detection and correction. The objective of this research study is to devise a way to improve the detection of such errors by using a cooperative approach, using metrics-based error detection. With the cooperation of 20 professional software developers, disciplined experiments were carried out, with preliminary results confirming their effectiveness in revealing design flaws. In addition, procedures were also produced in support of a cooperative approach to the problem, and the results can form the basis of further research and help individual software developers in their efforts to detect source code errors.