Deep Cross-Project Software Reliability Growth Model Using Project Similarity-Based Clustering

Kyawt Kyawt San,Akira Matsuzaki,Masahiro Taga,Hironori Washizaki,Kiyoshi Honda,Yoshiaki Fukazawa

doi:10.3390/math9222945

Abstract

Software reliability is an essential characteristic for ensuring the qualities of software products. Predicting the potential number of bugs from the beginning of a development project allows practitioners to make the appropriate decisions regarding testing activities. In the initial development phases, applying traditional software reliability growth models (SRGMs) with limited past data does not always provide reliable prediction result for decision making. To overcome this, herein, we propose a new software reliability modeling method called a deep cross-project software reliability growth model (DC-SRGM). DC-SRGM is a cross-project prediction method that uses features of previous projects’ data through project similarity. Specifically, the proposed method applies cluster-based project selection for the training data source and modeling by a deep learning method. Experiments involving 15 real datasets from a company and 11 open source software datasets show that DC-SRGM can more precisely describe the reliability of ongoing development projects than existing traditional SRGMs and the LSTM model.

Highlights

To adopt a more reliable cross-project method of software reliability growth modeling while eliminating the unrelated data from all source projects for each target project, this study introduces a new software reliability growth models (SRGMs) method which can be utilized at the beginning stage of ongoing projects by processing only the project data with the most common features of the target project
The industrial and open source software (OSS) datasets results indicated that DCSRGM outperformed long short-term memory (LSTM) and the Logistic model and improved the prediction accuracy when applied in an ongoing stage of industrial development
The case studies showed that deep cross-project software reliability growth model (DC-SRGM) is superior to all other evaluated models

Summary

Introduction

Reliability is one of the most significant attributes in enhancing the quality of the product in the software development process [1,2,3]. Growth Models (SRGMs) that are used for modeling the failure or defect arrival pattern [11]. Many SRGMs have been studied to measure the failure process. These models require external parameters to be estimated by the least-squares or maximum likelihood estimation to build the relevant parameters [1]. N. Ullah et al [11] studied different SRGMs using defect data in industrial and open source software and performed a comparative analysis between them. Honda et al [6]

Objectives

Methods

Results

Conclusion