A Reuse-oriented Clustering Method for Test Cases

Abstract

Different types of data are generated in each stage of software testing. There are a lot of test case data in the historical test asset library, including some case data with high similarity. Clustering test cases can effectively reduce the resource consumption and time cost of reusing test cases and improve the efficiency of test case recommendation. This paper proposes a reuse-oriented clustering method for test cases. Firstly, test case data of historical test projects of command-and-control system are collected, test case corpus is constructed, and word segmentation experiments are carried out using this corpus. Experimental tools are determined by comparing the experimental effects of current mainstream natural language word segmentation tools with self-defining dictionaries. Then, the keywords of test cases are extracted by keyword extraction algorithm, and the test case package is obtained by Spectral Clustering algorithm based on the test case similarity matrix and keyword similarity matrix. Finally, the validity of the proposed method is verified by experimental comparison on the constructed test case corpus.