Using Software Reliability Growth Models for Developing a Software Quality Index
Kashif Bilal, Hafiz Rizwan Rasheed · 2008
reliability growth modeling is a mathematical technique to predict and estimate the reliability of software using reliability growth models. Defect data pattern is comprehensively modeled with this technique, therefore it is widely gaining acceptance as measure for software quality. Our experiment utilizes most popular open source project repository for gathering project statistics. This is a two pronged study for sorting out the quality related statistics and estimating the reliability .Currently these repositories apply acceptance measures for aiding the user judgment about particular software these measures are relatively arbitrary and based on project statistics related to developer, user and traffic of a particular project. Unlike arbitrary formulation adopted by opensource repositories we establish a relationship between software reliability and pile of project statistics to finally extract out only those factors which have high impact on software quality .Using Computer aided software reliability engineering tool (CASRE) a methodology is proposed to estimate and predict the reliability of a software under observation and regression analysis is performed to elaborate the impact of developer, user, web traffic statistics on software reliability. We pointed out that developer activities related data highly impacts the software reliability. Sifting through the available statistics we aim at developing a software quality index which can serve as minimum information to effectively determine the quality of software.