Alternatives for facilitating automatic transformation of BIM data using semantic query languages

Abstract

In the Architecture Engineering and Construction (AEC) industry, Building Information Model (BIM) authoring tools enable the creation of digital representations of buildings. Each tool implements its own building data model, which makes it difficult to achieve the desired interoperability when building data have to be exchanged with other software (e.g., building energy simulation tools). The representation of BIM models through open standards (e.g., IFC) and Semantic Web technologies can facilitate building data transformation in an automated and flexible way. This is achieved by taking advantage of the logical basis of the Resource Description Framework (RDF) data model and queries created in the Semantic Web query languages. The result is a pragmatic mechanism to transform the data from one data domain to another. This article analyses the potential of Semantic Web query languages to facilitate the data transformation of building data through different alternatives. The first contribution is the identification of fourteen data mapping patterns and three cases of data transformation that enable transforming one data model into another, considering the semantic and structural differences between them. The second contribution is the review and comparison of query languages to carry out the transformations through two different alternatives: using SPARQL-Generate and SPARQL Construct queries. And finally, the third contribution is the definition of a metric to assess the complexity of SPARQL queries.