ANP-based software reliability prediction using PoCs and subsequent employment of orthogonal defect classification measurements for risk mitigation during prototype studies

Abstract

In a process centric software development, early reliability prediction plays a vital role in ensuring the project success. Inherent mitigation of risks can be obtained by relying on the conceptualization of Rational Unified Process (RUP). By implanting RUP methodologies into the project build, a continuum of benefits can be reaped intrinsically. Quality of Service (QoS) can be directly improved during the prototyping. Instead of having to strictly depend on prototype implementation for the necessary information that is vital to improvising upon the design functionalities, we have earlier developed frameworks that enable the bypassing of the prototyping modules. Now, a procedure to identify the potential critical phases of a software development process using Analytic Network Process (ANP), before the implementation of prototype is put forward. The results are compared with those obtained on realizing the prototype. While analyzing the information from the prototype realization, measures to strengthen the effectiveness of the whole set-up with the intent of risk mitigation are explained. This involves the reliance on Orthogonal Defect Classification (ODC) for a better reliability prediction.