Optimizing Software Requirements Selection: A Hybrid K-Means and Intuitionistic Fuzzy TOPSIS Approach

The success of a software development project relies greatly on the priority given to the software requirements. Such a process is however easy to capture because evaluation of this requirements is extremely prone to personal differences as well as risk. Our work offers a new contribution to the development of the software requirements prioritization process. This approach incorporates K-means clustering and intuitionistic fuzzy AHP. It classifies the similar requirements using the clustering technique, while the vague aspect of human perception is addressed using multi-criterion decision making MCDM. Practical aspects of the method are demonstrated with an illustrative example of requirement prioritization for an Examination System (ES). The results emphasize the potential of the approach in improving the transparency and accuracy of decision-making fluidity, in which the project managers can adopt it as a strategic tool for planning and resource distribution. Nevertheless, practical limitations been met, and reliance on expert judgement requires therefore further work to eliminate these constraints. All in all, this hybrid approach provides a strong basis for handling the challenges of prioritization of software requirements to provide better processes of software development that is responsive to changes.