Requirements prioritization plays a key role in the requirements engineering process, in particular with respect to critical tasks such as requirements negotiation and software release planning. This paper presents a novel framework which is based on a requirements prioritization process that interleaves human and machine activities, enabling for an accurate prioritization of requirements. Similarly to the Analytic Hierarchy Process (AHP) method, our framework adopts an elicitation process based on the acquisition of pairwise preferences. Differently from AHP, where scalability is a big issue, the framework enables a prioritization process even over a large set of requirements, thanks to the exploitation of machine learning techniques that induce requirements ranking approximations at run time, and to the use of a boolean metrics. Moreover the new approach allows to reduce the bias of a dominance hierarchy, a strategy introduced by AHP to deal with the scalability issue. The paper describes also a methodology for the experimental evaluation of the framework and discusses the results of a first set of experiments designed on a real case-study which shows that an high accuracy in the final ranking can be obtained within a limited elicitation effort

Supporting the Requirements Prioritization Process. A Machine Learning approach

Avesani, Paolo;Bazzanella, Cinzia;Perini, Anna;Susi, Angelo
2004

Abstract

Requirements prioritization plays a key role in the requirements engineering process, in particular with respect to critical tasks such as requirements negotiation and software release planning. This paper presents a novel framework which is based on a requirements prioritization process that interleaves human and machine activities, enabling for an accurate prioritization of requirements. Similarly to the Analytic Hierarchy Process (AHP) method, our framework adopts an elicitation process based on the acquisition of pairwise preferences. Differently from AHP, where scalability is a big issue, the framework enables a prioritization process even over a large set of requirements, thanks to the exploitation of machine learning techniques that induce requirements ranking approximations at run time, and to the use of a boolean metrics. Moreover the new approach allows to reduce the bias of a dominance hierarchy, a strategy introduced by AHP to deal with the scalability issue. The paper describes also a methodology for the experimental evaluation of the framework and discusses the results of a first set of experiments designed on a real case-study which shows that an high accuracy in the final ranking can be obtained within a limited elicitation effort
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: http://hdl.handle.net/11582/2524
 Attenzione

Attenzione! I dati visualizzati non sono stati sottoposti a validazione da parte dell'ateneo

Citazioni
  • ???jsp.display-item.citation.pmc??? ND
social impact