Exploring BIM Data by Graph-based Unsupervised Learning

This paper presents an unsupervised learning method for mining the Industry Foundation Classes (IFC) based Building Information Modelling (BIM) data by exploring the inter-relational graph-like building spaces. In our method, the affinity propagation clustering algorithm is adapted with our proposed feature extraction algorithm to get exemplars of certain spaces with similar usage functions. The experiments are conducted on a real world BIM dataset. The experimental results show that some build spaces of typical usage functions can be discovered by our unsupervised learning algorithm.

[1]  Alla Sheffer,et al.  Elements of style , 2015, ACM Trans. Graph..

[2]  Delbert Dueck,et al.  Clustering by Passing Messages Between Data Points , 2007, Science.

[3]  Tf Thomas Krijnen,et al.  Assessing implicit knowledge in BIM models with machine learning , 2015 .

[4]  Alla Sheffer,et al.  Elements of Style: Learning Perceptual Shape Style Similarity , 2015 .

[5]  Thomas A. Markus,et al.  The Words Between the Spaces: Buildings and Language , 2002 .

[6]  Gang Li,et al.  Recognition and evaluation of bridge cracks with modified active contour model and greedy search-based support vector machine , 2017 .

[7]  André Borrmann,et al.  Automated generation of building fingerprints using a spatio-semantic query language for building information models , 2014 .

[8]  B. Bjoerk.  INDUSTRY FOUNDATION CLASSES FOR PROJECT MANAGEMENT - A TRIAL IMPLEMENTATION , 1999 .

[9]  C. Soddu.  Generative designer, a swimmer in a natural sea frame , 2006 .

[10]  Thomas A. Markus.  The Social Logic of Space, Bill Hillier, Julienne Hanson. Cambridge University Press, Chicago (1984), 281 pp. £35.00 , 1986 .

[11]  Esko Sistonen,et al.  Machine learning for durability and service-life assessment of reinforced concrete structures: Recent advances and future directions , 2017 .

[12]  Bill Hillier,et al.  Space is the machine: A configurational theory of architecture , 1996 .

[13]  Lutz Plümer,et al.  Automatic classification of building types in 3D city models , 2012, GeoInformatica.

[14]  Hau Yan Leung,et al.  Building Information Modelling (BIM): A new paradigm for visual interactive modeling and simulation for construction project , 2008 .