Title: Software change impact analysis: an approach to differentiate type of change to minimise regression test selection

Authors: Chetna Gupta; Maneesha Srivastav; Varun Gupta

Addresses: Department of CSE, JIIT Noida 201307, India ' Department of CSE, JIIT Noida 201307, India ' Department of Computer Science and Engineering, Jawaharlal Nehru Govt. Engineering College, Sundernagar, Himachal Pradesh 175001, India

Abstract: Software evolution is a continuous process carried out with the aim of extending base applications either for adding new functionalities or for adapting software to the changing environment. This paper proposes a new approach of estimating impact analysis by allocating tokens to the changes encountered in the two versions of the software system. An algorithm is proposed for token allocation and for determining matches which take into account a minimum threshold value to predict the matched results. To establish confidence in effectiveness and efficiency, presented technique is illustrated with the help of an example and the results of analysis are promising towards achieving the aim of the proposed endeavour. Further, a performance-based comparison between existing techniques is also provided in support of this research. The impact set produced will be more precise than other techniques and this data can then be used by a software engineer in determining the changes made to the software system. Thus, the proposed technique can help software engineer to perform selective regression testing by optimising the number of test cases.

Keywords: change impact analysis; regression testing; software maintenance; software testing; software changes; regression test selection; software development; token allocation.

DOI: 10.1504/IJCAT.2015.070498

International Journal of Computer Applications in Technology, 2015 Vol.51 No.4, pp.366 - 375

Published online: 08 Jul 2015 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article