IsingBench

Test optimization problems such as test case selection and test case minimization are long-standing combinatorial optimization challenges in software testing. Ising-based computing offers a new way to formulate and solve such problems by representing them through Ising models and executing them with different solver paradigms.

IsingBench provides an open-source environment for studying this direction systematically. It brings together benchmark datasets, Ising formulations, configurable solvers, and result analysis in a unified workflow, enabling researchers to evaluate Ising-based test optimization methods under consistent experimental settings.

The framework is accompanied by IsingTester, which supports the end-to-end process of encoding test optimization problems into Ising representations, executing selected solvers, and decoding the resulting solutions back into optimized test suites.