Abstract:
"
Regression testing is the method of retesting software that has been updated or extended during
software development. Usually, the execution of a series of test cases is used to conduct
regression testing. However, it can be expensive and time-consuming to keep the regression
suite up to date. Furthermore, executing all of the test cases in a regression suite could take a
long time. Quality engineers in the industry are constantly confronted with this problem, and
they must find bugs as soon as possible.
To overcome this issue, test case prioritization is one technique that is used to execute
regression testing efficiently. Test case prioritization can greatly improve regression testing
through organizing test cases according to their goals and optimizing their execution. Although
there are numerous methods for prioritizing test cases, they all need knowledge of the internal
source code (white-box testing). In industry, black box testing is an important requirement that
typically does not allow access to source code.
This research project introduces a novel supervised machine learning and natural language
processing-based technique in order to prioritizing the test cases which is currently a manual
process of most industrial companies. This proposed system prioritizes the test cases into
several priority levels as High, Medium and Low, which will be a perfect companion for quality
engineers to make their life easier.
"
