A Probabilistic Model for Analysis and Fault Detection in the Software System: An Empirical Approach

Abstract

Software reliability and estimation of defects plays an important role in software testing stage. For studying defects, one common practice is to inject faults in subject software, either manually or by using a program that generates all possible mutants based on a set of mutation constraints. Getting the optimized results for the software system while predicting defects using realistic analysis, and confirming whether that leads to valid and consistent data during software testing stage is a challenge. In this paper, we propose Process simulation Model (PSM), which is a probabilistic model-based approach that overcomes these challenges and enables prediction of software defects and its impact in the system using Bayesian estimation. Moreover, a Fault Detection Algorithm FDA is derived from PSM model that helps to predict software faults for different deterministic problems that we have taken in our experimental study to demonstrate the reliability, verification and consistency of the system. A comparative study is shown on various deterministic problems by finding set of random defects through probabilistic approach the fault may occur in the proposed software model.