Looking for conflicts in software policies is a critical activity in many contexts, like security, expert systems, and databases. These policies are often written or specified as logical rule-based systems, which is a flexible and modular formalism. However, this formalism does not guarantee by itself the safety of the system. While several works address conflict detection, as far as we know there is little research on making explicit all these problems in an abstract and simplified form. In this paper, we compare different techniques to extract all the problems in a rule-based system. We define an exact algorithm and an application on three middle-size case studies, but as expected with poor time performances. We improve this with a combine algorithm and a few simple heuristics that accelerate the computation of all the problems in the case studies by several orders of magnitude.