EXPLORING BIM AND NLP APPLICATIONS: A SCIENTOMETRIC APPROACH

Abstract

In AECO (Architecture, Engineering, Construction and Owner-operated) industry information are mainly defined and exchanged in natural language through textual files and documents. On the contrary, a Building Information Modeling (BIM) approach requires machine processable data and information. Natural Language Processing (NLP) allows to process textual information into structured format. About BIM in AECO industry, and NLP in different sectors, several literature reviews have been conducted. However, none of them highlighted the possible connections between the two topics. This study provides a scientometric analysis aiming to investigate possible combined applications of BIM and NLP. A quantitative literature review approach is employed, using data visualization and science mapping applied on bibliometric meta-data. The performed analysis uncovered possible directions for further research on NLP and BIM combined applications in AECO sector. Most active authors, key research patterns, and institutional affiliations are identified. The main applications areas of a combined NLP and BIM approach in AECO are: Information Retrieval and Information Enrichment of BIM models, Automatic Compliance Checking, and Safety and Risk Management. The keywords pattern analysis highlighted the main tools which allows to link Semantic BIM and NLP methods and technologies, i.e. Ontology and Machine Learning algorithms. The scientometric analysis also reveals a gap related to the Preliminary design and Requirement definition phases, highlighting a possible research area not covered by the Academia as of now.