FREQUENT ITEM TEST CASE CLUSTERING BASED TEST SUITE REDUCTION

摘要

Execution profiles are indicators for code coverage of program; this has been demonstrated by researchers on a large scale through their contributions on the same. Test Suite reduction is a feature which achieves code coverage with minimum number of test cases ensuring that all code items have been tested. It is a NP-Complete problem. Few approaches like Greedy approach, HGS approach have been proposed in literature which are good approaches. Current work achieves similar milestones with reduced test cases. This paper presents Maximal frequency item set clustering for test cases, residue code requirements based test case reduction and modification based test case selection. In the current work, few interesting results were found, where in similar program trace test cases were greatly reduced ensuring high code coverage percentage in initial stages during testing. Fault detection can be tuned by selecting test cases from similar groups.