International Journal of Computer Applications |
Foundation of Computer Science (FCS), NY, USA |
Volume 78 - Number 4 |
Year of Publication: 2013 |
Authors: Priya Purohit |
10.5120/13480-1174 |
Priya Purohit . An Automated Sequence Model Testing (ASMT) for Reducing Test Suit Size. International Journal of Computer Applications. 78, 4 ( September 2013), 32-37. DOI=10.5120/13480-1174
Quality assertion of software is primarily done by means of testing an activity that faces constraints of both time & resources. Conventional testing strategies required the heavy resources & it can be done after the complete code is developed. So to apply some changes after the development requires lots of effort & cost. It also affects the time of deployment. So model based testing is a well accepted & dynamic approach for quality improvements because it provides affecting error detection at very low cost. It gives scalable & systematic solution to the test case reduction problem. To better understand the need of this early test case generation overall scenario of collaboration & development strategies is to be considered. The best way to get better knowledge of internal structure design models is the best option. Thus in this work we are focusing our research on path based testing by which internal structure of complete design can be verified & tested. Model based testing on path coverage is applied with standard coverage criteria for the large software's & generates larger test counts. Our aim is to reduce those test counts with maximum coverage of testing at the early phases of SDLC even before the actual development starts. This paper proposes a novel automated sequence model based testing (ASMT) approach selects the test criteria based on UML diagram like activity, sequence, class etc. It applies the criteria with generation modules to create novel test cases. The typical deployment of ASMT goes through five stages: setting up test criteria, test model designing, test suite creation, performing test & analyzing the result. Thus by combining all the above methods improved test strategies can be designed. At the initial level of our research the approach seems to be better that others & will proves its effectiveness in future implementation.