A comprehensive study for software testing and test cases generation paradigms

Abstract

Software testing is accounted to be an essential part in software development life cycle in terms of cost and manpower, where its total cost is considerable high. Consequently, many studies [48] have been conducted to minimize the associated cost and human effort to fix bugs and errors, and to improve the quality of testing process by automatically generating test cases. Test cases can be generated from different phases (requirement phase, design phase and after development phase). Though, test case generation at early stages is more effective rather than that after development, where time and effort used for finding and fixing errors and bugs is less than that after development. At later stage, fixing errors results in enormous code correction, consuming a lot of time and effort. In this paper, we study the different paradigms of testing techniques for generating test cases, where we investigate their coverage and associated capabilities. We finally propose a preliminary model for a generic automated test cases generation.