Multi-scale Information Retrieval for BIM using Hierarchical Structure Modelling and Natural Language Processing

Abstract

Building Information Modelling (BIM) captures numerous information the life cycle of buildings. Information retrieval is one of fundamental tasks for BIM decision support systems. Currently, most of the BIM retrieval systems focused on querying existing BIM models from a BIM database, seldom studies explore the multi-scale information retrieval from a BIM model. This study proposes a multi-scale information retrieval scheme for BIM jointly using the hierarchical structure of BIM and Natural Language Processing (NLP). Firstly, a BIM Hierarchy Tree (BIH-Tree) model is constructed to interpret the hierarchical structure relations among BIM data according to Industry Foundation Class (IFC) specification. Secondly, technologies of NLP and International Framework for Dictionaries (IFD) are employed to parse and unify the queries. Thirdly, a novel information retrieval scheme is developed to find the multi-scale information associated with the unified queries. Finally, the retrieval method proposed in this study is applied to an engineering case, and the practical results show that the proposed method is effective.

Automation of text document classification in the budgeting phase of the Construction process: a Systematic Literature Review

“…It manifests itself as an impediment to selecting the sample adequately, possibly affecting the study's results (Pannucci and Wilkins, 2010). For instance, in Wang et al(2021b), all the information comes from a BIM model, which can affect the results in terms of accuracy and error. Also, (Akanbi and Zhang, 2020;Dimitriou et al, 2018) use only eight projects of a single scope (wood construction and CI 24,7 road projects, respectively).…”

“…Most of the work studied started with digital text files, which were then classified. However, there is also a more ambitious approach starting from BIM models (Guo et al, 2021;Juszczyk, 2018aJuszczyk, , 2018bWang et al, 2021b).…”

Demystifying the Role of Natural Language Processing (NLP) in Smart City Applications: Background, Motivation, Recent Advances, and Future Research Directions

“…Bommarito et al [125] introduced LexNLP for the task of IE in legal and regulatory text documents. A multi-scale IR model that uses NLP besides a hierarchy tree structure was proposed by Wang et al [126] for the task of Building Information Modelling (BIM).…”

Algoritmos de classificação de texto na automatização dos processos orçamentação

“…Um outro método para obtenção de dados é a utilização de modelos Building Information Modelling (BIM) [10 -13]. Um modelo BIM capta muita informação sobre o ciclo de vida dos edifícios, sendo que é através da recuperação desta informação que os sistemas de apoio à decisão utilizam o BIM [13]. Wang [13], propõe um esquema de recuperação de informação para BIM utilizando, para isso, uma estrutura de árvore hierárquica e técnicas NLP.…”

“…Um modelo BIM capta muita informação sobre o ciclo de vida dos edifícios, sendo que é através da recuperação desta informação que os sistemas de apoio à decisão utilizam o BIM [13]. Wang [13], propõe um esquema de recuperação de informação para BIM utilizando, para isso, uma estrutura de árvore hierárquica e técnicas NLP. Através deste sistema o autor consegue aceder à informação dentro do modelo BIM através da estrutura de dados da especificação Industry Foundation Class (IFC), analisando esta informação com a ajuda do International Framework for Dictionaries (IFD) e algoritmos NLP para desenvolver um novo esquema de recuperação de informação para detetar informação multiforme associada às consultas de modelos BIM.…”