Automated software test data generation using improved search procedure

Abstract

Automated software test data generation is a complex and one of the most challenging task. Generation of test data manually is tedious and error-prone. The base algorithm, which has been improved in this paper involves redundant and un-necessary execution cycles. In this paper, we have proposed an efficient search procedure that reduces the time complexity of test data generation significantly. Our experiments conducted on varying sizes of source code showing ratio of execution time of automated test data generation using Basic Search Procedure and our Improved Search Procedure conclude that, with the increase of source code size, test data generation takes place in lesser time using our improved search procedure.