RUB » Bibliotheksportal

Kilian Jonathan Kandt, Sven Zentgraf

• Most of the information relating to the requirements of construction practices and buildings can be found in standards and regulations. However, most of these requirements are not available in a machine-readable format. Some approaches such as the Initiative Smart Standards of the German Institute for Standardization (DIN) aim to address this issue by presenting the data as NISO Standards Tag Suite (NISO-STS), an XML model for publishing and exchanging full-text content and metadata of standards. The DIN EN ISO 23386 is a civil engineering standard that proposes a methodology for describing building element properties as interconnected networks via property groups. The IDDO ontology implements the data model of ISO 23386. To enhance the properties and groups with textual information, the NISO-STS files are utilized. To extract this data, the textual content within the NISO-STS XML documents needs to be scraped and the extracted results must be validated. This paper presents an approach to extracting textual information from NISO-STS XML files to enrich the IDDO-Ontology properties and property groups. To illustrate the applicability of the Python-based crawler, the extracted textual information is further processed using different Natural Language Processing (NLP) algorithms.