Emunds Christoph ; Pauen Nicolas ; Richter Veronika ; Frisch Jérôme ; van Treeck Christoph

Abstract

Enhancing interoperability and information exchange between domain-specificsoftware products for BIM is an important aspect in the Architecture,Engineering, Construction and Operations industry. Recent research startedinvestigating methods from the areas of machine and deep learning for semanticenrichment of BIM models. However, training and evaluation of these machinelearning algorithms requires sufficiently large and comprehensive datasets.This work presents IFCNet, a dataset of single-entity IFC files spanning abroad range of IFC classes containing both geometric and semantic information.Using only the geometric information of objects, the experiments show thatthree different deep learning models are able to achieve good classificationperformance.