Title: Potential-Fault Cache-Based Regression Test Selection
Author : Samia El-Sayed Megahed
Collection : M.Sc. Computer
Abstract:
Early prediction of bugs decreases the cost and the effort required for bug fixes and improves the quality and reliability of software projects. Especially for the local Small and Medium Entreprises (SMEs) that behold the main contribution in the software industry.
Early detection of bugs improves the quality of software products by decreasing the time needed for maintenance. Running test suites frequently enforces early detection of problems in a product. However, due to limited resources of SMEs, test suites are often expensive to execute frequently. Cache-based techniques presented early bug detection methods that depend on a cache-like structure to hold the set of buggy classes throughout the different development stages.
The cache-based techniques have been extended to be used in regression test selection.
In this thesis, a regression test selection technique is proposed. The proposed technique adopts the idea of the cache-based techniques in regression test selection and utilizes the software repositories data to enhance the cache operations and test cases selection method.
The most error prone files are identified using temporal locality and static code analysis metrics and a classifier is constructed to help in assessing the fault proneness of files.
ليست هناك تعليقات:
إرسال تعليق