On the basis of a systematic literature review, this thesis summarizes state-of-the-art utilization of LCA, BIM, and NLP in the AEC industry, and compares three kinds of BIM-LCA integration strategies: manual, semi-automated, and automated approaches. Moreover, an automated method of mapping LCA data to BIM models which introduces NLP technology is proposed, including the methodology and a prototypical workflow. To verify the feasibility of this method, a case study is implemented, where an IFC file is at first extracted from a BIM model, followed by the processing of the LCA-related data contained in the IFC file. Then the processed information (material names) is mapped to the LCA database Ökobaudat respectively through manual attempt and three different NLP tools – Gensim, spaCy, and BERT. In the end, after quantitatively analyzing and comparing the contributions of the three NLP technologies to the accuracy of the mapping task, performing the automated method based on the pre-trained BERT model in conjunction with manual checking and adjustment is concluded to be the most efficient and recommendable.     «

On the basis of a systematic literature review, this thesis summarizes state-of-the-art utilization of LCA, BIM, and NLP in the AEC industry, and compares three kinds of BIM-LCA integration strategies: manual, semi-automated, and automated approaches. Moreover, an automated method of mapping LCA data to BIM models which introduces NLP technology is proposed, including the methodology and a prototypical workflow. To verify the feasibility of this method, a case study is implemented, where an IFC...     »