Natural Language Processing and Bim in Aeco Sector: A State of the Art

Abstract

The research provides state-of-the-art theories, methods, and applications of Natural Language Processing in a BIM approach to define advantages, weaknesses, and potential developments. Traditionally, the design and construction process requirements are expressed through verbal qualitative evaluations instead of using numerical and structured data. Information modeling and management methods can hardly manage requirements expressed by means of qualitative and unstructured expressions. This paper aims to define the state of the art on Natural Language Processing applications for the numerical translation of basic design requirements in the AECO sector. The major advantages of this research would be the optimization and automation of numerical definition and validation of requirements in a structured data form. Structured data can, in fact, be easily managed in a BIM approach. NLP supporting information modeling methods allows the application of requirements engineering and management techniques to handle construction processes from a data-driven perspective. An investigation of the setting of a decision support system, based on NLP, for the definition and validation of requirements in the early stages of the construction process, is also provided as a potential further development.

Fostering the consensus: a BERT-based Multi-label Text Classifier to support agreement in public design call for tenders

“…Project, Safety, and Risk Management are the areas with the highest number of applications. Moreover, cases of combined applications of the Building Information Modelling (BIM) method and text processing to streamline Automated Compliance Checking tasks are currently being investigated, and NLP-based systems to convert regulatory information represent an active field of research [9]. However, in the scientific literature, no applications are related to the pre-design phase and specifically to the processing of quality objectives and needs expressed in written form via text documents [10].…”