Improving fault detection rate using similarity-based test case prioritization in regression testing

Abstract

Rapid evolution in software requires regression testing to be performed as an essential activity which validates the software before the next release. Where software developer may add or removes intended features to maintain the software according to the customer requirements. In that case, complete test cases execution is nearly infeasible due to limited time and resources. So, the main aim of prioritization is to test any software with minimal time and maximum efficiency in terms of fault coverage rate. This paper proposed different similarity-based prioritization techniques to provide ranking to the test cases based on their influence level which is computed as similarity degree in three levels for the software to be tested. Each level represents the integration of selected coverage criteria’s. In order to validate our proposed technique, we have conducted a case study to measure its effectiveness in prioritizing the test cases. We experimentally observed that by incorporating a similarity-based approach with more than one coverage criteria; results for similarity-based prioritization are promising than any other conventional coverage based approaches in terms of Average Percentage of Faults Detected.