Business experts of building are overwhelmed by regulatory texts. It is a heavy task to go through these texts and get an unambiguous list of requirements they contain. More, with regard to the number of texts and the diversity of their writers, we cannot omit the possibility of getting inconsistencies. Finally, these requirements are to be put closed to digital representation of buildings to detect potential non-conformities. This PhD thesis examines these problems and envisions solutions to help experts. We thus envisage to automate detection, extraction and annotation of business rules in regulatory texts. Next to formalise identified requirements as SPARQL queries. These queries will serve for conformity checking on OWL-representation of buildings. Moreover, we plan to leverage these queries to detect inconsistencies in texts. Further, all these tasks will be accessed through business services.