Performance Evaluation of Clustering Techniques in Test Case Prioritization

Regression testing plays a crucial role in maintaining quality of a software, yet accounts for a huge percentage of cost from overall development cost. Selection of regression testing technique directly impacts the software quality, where at first step relevant test cases are selected, then redundant test case are removed during minimization step and at final step test cases are prioritized to execute the most relevant test cases first and so on. The test case prioritization is one of the broadly used approach to reduce cost and time of regression testing. In literature researchers have proposed various methods to prioritize test cases, and clustering is one of the popular and suggested techniques among them. It is an unsupervised method of putting similar data into one cluster and dissimilar data into different cluster and considered to be an important tool for exploratory data analysis. This paper analyses various clustering techniques used for test case prioritization and presents a performance analysis on different hard clustering algorithm and then the test case prioritization techniques are also evaluated using APFD.