Test Case Reduction Case Study for White Box Testing and Black Box Testing using Data Mining

The work demonstrates the case study of using black box testing and white box testing. It attempts to reduce the size of test cases generated. The data mining approach is explored. We present the methodology starting from the test case generation for the block box and the white box approaches. Pex software[1] is used to generate test cases for white box testing. Also, for the black box testing, we analyze input variables for each condition. We use statistical values to analyze related variables and select variables. Then, we use clustering to reduce test cases. We can further reduce the test case from 648 to 486 and to 324 for black box testing using K-Means while clustering keeping the same coverage. For the white box testing, Pex already generates the small size test case. Thus, it is not meaningful to reduce further