Employing machine learning techniques to assess requirement change volatility

Abstract Lack of planning when changing requirements to reflect stakeholders’ expectations can lead propagated changes that cause project failures. Existing tools cannot provide the formal reasoning required manage requirement change and minimize unanticipated propagation. This research explores machine learning techniques predict volatility (RCV) using complex network metrics based on premise networks be utilized study Three questions (RQs) are addressed: (1) Can RCV measured through four classes namely, multiplier, absorber, transmitter, robust, during every instance change? (2) explored computed for each (3) techniques, specifically, multilabel (MLL) methods employed metrics? in this paper quantifies propagation, is, how behave response initial change. A multiplier is a changed by an propagates other requirements. An absorber change, but does not propagate transmitter robust determined industrial data relationships obtained from previously developed Refined Automated Requirement Change Propagation Prediction (R-ARCPP) tool. Useful highest performing models discussed along with limitations future directions research.