Automating test case generation helps ensure fully functional, accurate application development. Those defects in software cause applications to fail and web sites to have to be taken down. Ideally defects should be detected and fixed before the application goes live but it's the result when software is not tested properly. Philip Howard with Bloor Research warns that manual software testing results in more defects and recommends that test cases are both tied to requirements and automatically generated.
This paper explores not the testing per se but rather on automating the testing process.