APPLICATION OF NEURAL NETWORKS TO SOFTWARE QUALITY MODELING OF A VERYLARGE TELECOMMUNICATIONS SYSTEM

Society relies on telecommunications to such an extent that telecommun ications software must have high reliability. Enhanced measurement for early risk assessment of latent defects (EMERALD) is a joint project of Nortel and Bell Canada for improving the reliability of telecommuni cations software products. This paper reports a case study of neural-n etwork modeling techniques developed for the EMERALD system. The resul ting neural network is currently in the prototype testing phase at Nor tel. Neural-network models can be used to identify fault-prone modules for extra attention early in development, and thus reduce the risk of operational problems with those modules. We modeled a subset of modul es representing over seven million lines of code from a very large tel ecommunications software system. The set consisted of those modules re used with changes from the previous release. The dependent variable wa s membership in the class of fault-prone modules. The independent vari ables were principal components of nine measures of software design at tributes. We compared the neural-network model with a nonparametric di scriminant model and found the neural-network model had better predict ive accuracy.