Software Product Quality Assessment Model Using Deep Learning

Abstract

Software quality assessment model is a metric based approach for evaluating and monitor quality of software product and all quality assessment models have a clear methods on how to assess a software system.Assessing quality of software product using Convolutional neural network modelshelps to understand how to assess maintainability,reliability,portability and usability of software products and used to bridge the gap between quality metrics and software quality factors with the help of quality assessment model.Quality assessment models assess the quality of software under investigation by determining the values of quality factors using aggregation methods.In this studyOne dimentional Convolutional neural network algorithm has been implemented to build deep learning software product quality assessment model with regard to its computational capacities to evaluate values of software product quality attributes.This study is aimed to determine a useful way to evaluate a values of maintainability,reliability,portability and usability quality factors by aggregating software product internal quality metrics using deep learning methods.Datasets taken from PROMISE software engineering repository had been used for training and testing the model.We developed Convolutional neural network model that helps to evaluate quality of software product.as a result of implementing designed model to datasets,result shows that accuracy of 97% had been achieved.Consequently,we found that our model can be used to evaluate and assessquality of any software product using software product quality assessment datasets.