Extending Randomness: Towards Uniformity for Cardinality-Based Feature Model Sampling

  • Type:Bachelor's thesis
  • Supervisor:

    Lukas Güthing

  • Person in Charge:Open

  • Context: Cardinality-based Feature Models (CFMs) extend classical boolean feature models with multiplicities of features. While there are Uniform Random Sampling algorithms for sampling boolean FMs in order to enable sample-based testing, random sampling for CFMs is only naively done, without ensuring uniformity.

     

    Goal: Find, implement, and possibly compare heuristics to improve uniformity of random sampling for CFMs.

     

    Requirements: Prior knowledge of product lines and solvers (CSP/SMT/ILP) is not required but might be helpful. Interest in combinatorics is probably useful.