Automating Inference of OCL Business Rules from User Scenarios

International audienceUser Scenarios have been advocated as an effective means to capture requirements by describing the system-to-be at the instance or example level. This instance-level information is then used to infer a possible software specification consistent with the provided valid and invalid scenarios. So far existing approaches have often focused on the generation of static models but have omitted the inference of business rules that could complement the static models and improve the precision of the software specification. In this sense this paper provides a first set of invariant inference patterns that are applied on valid and invalid snapshots in order to generate OCL~(Object Constraint Language) integrity constraints that the system should always satisfy. We strengthen the confidence of inferred results based on the user's feedback of generated examples and counterexamples for the considered constraint. The approach is realized with a prolog-based tool that could support the designer to effectively define OCL integrity constraints in a semi-automatic way