An induction-based method for retrieving similar cases and/or easily adaptable cases is presented in a 3-steps process: first, a rule set is learned from a data set; second, a reformulation of the problem domain is derived from this ruleset; third, a surface similarity with respect to the reformulated problem appears to be a structural similarity with respect to the initial representation of the domain. This method achieves some integration between machine learning and case-based reasoning: it uses both compiled knowledge (through the similarity measure and the ruleset it is derived from) and instanciated knowledge (through the cases).