Combinatorial testing is a black-box testing approach that derives test cases from a model defined by the tester. This approach generates test inputs such that each combination, of n values from different parameters, is at least covered by one test case. Thereby, the size n of the combinations can be defined by the tester and has to be greater than one. Once the test cases have been executed the tester is left with a possible list of failing input combinations. Finding the interaction of values leading to the error becomes with increasing number of test cases harder and harder. In order to support support the tester to find the minimal subset of values leading to an error fault characterization techniques have been developed. Additionally, tools have been developed that define the model independently of the programming language of the system under test. Therefore, those tools can be used for every programming language instead of a specific one. Those tools only automate the generation of the test inputs and do not support any fault characterization. Hence, this thesis provides a concept for a combinatorial testing framework that automates the combinatorial testing process and provides a web interface that allows the user to define the model programming language independently through a domain-specific language. As a proof of concept, the framework has been implemented and evaluated. Finally, suggestions for future research are given.