Daily Build Strategies using Combinatorial Interaction Testing

In a daily build process (which is also often referred to as a nightly build process),  a system under test (SUT) is automatically tested every day (typically during night times when the developers are away from their desks) by getting the latest version of the SUT from a code repository, such as git or SVN, building it, and running a predetermined set of test cases. The ultimate goal is to reveal the defects as soon as possible, since the more the time gap between introducing a defect in a code base and becoming aware of it, the more expensive typically it is to fix the defect.
Combinatorial Interaction Testing (CIT) is a well-known approach for systematically sampling combinatorial spaces for testing. CIT typically models a system under test (SUT) as a set of factors (choice points or parameters), each of which takes its values from a particular domain. Based on this model, CIT then generates a sample that meets specified coverage criteria; that is, the sample contains specified combinations of the factors and their values. For instance, pairwise testing requires that each possible combination of values, for each pair of factors, appears at least once in the sample. 
The goal of this project is to 1) develop CIT-based daily build strategies and 2) evaluate these strategies by conducting experiments on real-life software systems.