Combining genetic algorithm and pairwise testing for optimised test generation from UML ADs. Issue 5 (1st October 2019)

• Record Type: Journal Article
• Authors: R, Anbunathan; Basu, Anirban
• Is Part Of: IET software. Volume 13:Issue 5(2019)
• Page Start: 423
• Page End: 433
• Publication Date: 2019-10-01
• Subjects: genetic algorithms -- program testing -- Unified Modeling Language; genetic algorithm -- pairwise testing -- optimised test generation -- UML activity diagram -- test cases -- test data -- concurrent activities; Computer software -- Periodicals; Software engineering -- Periodicals; 005.1
• DOI: 10.1049/iet-sen.2018.5207
• Languages: English
• ISSNs: 1751-8806

Abstract

Abstract : Unified modelling language (UML) activity diagram (AD) is used to represent system behaviour abstractly and is used by testers to generate test cases and test data. During the design of test cases, AD with concurrent activities may lead to a large number of paths, and it may not always be possible to test all execution paths. Research on deriving test cases from AD having concurrent activities has focused on conventional search techniques such as breadth‐first search and depth‐first search which has been found to be inefficient in such cases. To overcome this drawback, the authors propose a method using pairwise testing and genetic algorithm to derive a reduced number of test cases in AD with concurrent activities. Experiments conducted on various real‐life concurrent systems show that the proposed technique generates a reduced number of test cases compared with existing methods.