Résumé : With the arrival of digital technologies in the field of architectural documentation, many tools and methods for data acquisition have been considerably developed. However, these developments are primarily used for recording colorimetric and dimensional properties of the objects processed. The actors, of the disciplines concerned by 3D digitization of architectural heritage, are facing with a large number of data, leaving the survey far from its cognitive dimension. In this context, it seems necessary to provide innovative solutions in order to increase the informational value of the representations produced by strengthen relations between "multiplicity" of data and "intelligibility" of the theoretical model. With the purpose of answering to the lack of methodology we perceived, this article therefore offers an approach to the creation of representation systems that articulate the digital instance with the geometric/semantic model.