ABSTRACT. This paper describes an application worked out in collaboration with a company that produces made-to-order machine components. The goal of the project is to develop a system that can support the company's engineers by automating parts of the component design process. We propose a knowledge extraction methodology based on the recent DMN (Decision Modelling Notation) standard and compare a rule-based and a constraint-based method for representing the resulting knowledge. We study the advantages and disadvantages of both approaches in the context of the company's real-life application.