pbourreau
commented
7 years ago Work In progress. Many aspects identified:
- parsing as quite simple. Many tools are available: https://github.com/mvaerle/python-ifc
- converting IFC 2 RDF is a tedious work. We can either reuse the work of Pauwel https://github.com/IDLabResearch/IFC-to-RDF-converter/ (but it is a Java program) or write our own Python converter. First option is considered.
- we have to be careful as we will probably not extract too much information from the IFC (only a short building description). Also, ifcOwl defines relation as concepts... Can this be simplified in our data model?
Develop a Python parser for the IFCs that are available in the project - as far as we know only NUIG, IFC 2X3_TC1. The parser will: -extract only relevant information (for instance, no spatial/geometric information) -create RDF triples from this information according to the schemas in IFC-OWL: http://ifcowl.openbimstandards.org/IFC2X3_TC1/index.html -insert the data into a Jena instance.