Using Interactive GA for Requirements Prioritization

Abstract

The order in which requirements are implemented in a system affects the value delivered to the final users in the successive releases of the system. Requirements prioritization aims at ranking the requirements so as to trade off user priorities and implementation constraints, such as technical dependencies among requirements and necessarily limited resources allocated to the project. Requirement analysts possess relevant knowledge about the relative importance of requirements. We use an Interactive Genetic Algorithm to produce a requirement ordering which complies with the existing priorities, satisfies the technical constraints and takes into account the relative preferences elicited from the user. On a real case study, we show that this approach improves non interactive optimization, ignoring the elicited preferences, and that it can handle a number of requirements which is otherwise problematic for state of the art techniques.