Test case prioritisation during web application testing

Abstract

Owing to frequent alterations in the existing web applications, performing regression testing becomes necessary for the identification and rectification of the newly generated unwanted faults. Owing to various resource constraints, test case prioritisation is one of the strategies followed, rather than running test cases blindly. This paper proposes a novel approach towards prioritisation of test cases during regression testing of web application (dynamic website) using Bayesian network. Initially, a Bayesian network is formed using various parameters which affect the success of a test case as well as promote testing of more crucial sections of the web application. Thereafter, the conditional probability table and probabilistic inference algorithms are applied to evaluate the success probability and ultimately priority (importance) of a test case. Execution of the test cases takes place on the basis of their respective priority. The performance of proposed technique is compared with existing work, 2-opt inspired heuristic algorithm and genetic algorithm.