Additive manufacturing (AM) or 3D printing, as an enabling technology for mass customisation or personalization, has been developed rapidly in recent years. Va r io u s design tools, materials, machines and service bureaus can be found in the market. Clearly, the choices are abundant, but users can be easily confused as to which AM process they should use. This paper first reviews the existing multi-attribute decision-making methods for AM process selection and assesses their suitability with regards to two aspects, preference rating flexibility and performance evaluation objectivity. We propose that an approach that is capable of handling incomplete attribute information and objective assessment within inherent data has advantages over other approaches. Based on this proposition, this paper proposes a weighted preference graph method for personalized preference evaluation, and a rough set based fuzzy axiomatic design approach for performance evaluation and the selection of appropriate AM processes.An example based on the previous research work of AM machine selection is given to validate its robustness for the priori articulation of AM process selection decision support.