CDBR: A semi-automated collaborative execute-before-after dependency-based requirement prioritization approach

Abstract

The success of requirement prioritization process largely depends upon how well different constraints and influential factors are handled by stakeholders and developers while prioritization. The main goal of this research is to present a semi-automated dependency based collaborative requirement prioritization approach (CDBR), which uses linguistic values, execute-before-after (EBA) relation among requirements and machine learning algorithm to minimize the difference of opinion between stakeholder and developers for effective collaboration and for better approximation of final prioritization results, acceptable to both. The presented approach targets three major constraints rarely addressed in existing work, namely dependencies among requirements, communication among stakeholder and developers and the issue of scalability. Results of performance assessment conducted on several different requirement sets and on a case study by comparing CDBR with other state of the art approaches namely, AHP and IGA. The results are accurate and comparable in terms of effectiveness, efficiency, scalability and disagreement concerns among stakeholder and developers which in turn provides robustness to decision making process of awarding more importance to some requirements over others. CDBR overpowers AHP and IGA in terms of efficiency and processing time respectively.