Applying Assemble Clustering Algorithm and Fault Prediction to Test Case Prioritization

Abstract

Cluster application is proposed as an efficient approach to improve test case prioritization, Test case in a same cluster are considered to have similar behaviors. In the process of cluster test case, the selection of test case feature and the clusters number have great influence on the clustering results. but to date almost clustering algorithm to improve test case prioritization are selected random clusters number and clustering result are based on one or a few of the code features, the paper propose a new prioritization techniques that not only consider the best clusters number but also produce the best clustering result based on test case multidimensional feature. After clustering, considering the inter-cluster prioritization and intra-cluster prioritization,in order to improve the effectiveness of our approach, the fault prediction value of code corresponding to the test case is used as one of a prioritization weight. Finally,we implemented an empirical studies using an industrial software to illustrate the effectiveness of the test case prioritization techniques.