NLP and Deep Learning-based Analysis of Building Regulations to Support Automated Rule Checking System

This paper aims to describe a natural language processing (NLP) and deep learning-based approach for supporting automated rule checking system. Automated rule checking has been developed in various ways and enhanced the efficiency of building design review process. Converting human-readable building regulations to computer-readable format is, however, still time-consuming and error-prone due to the nature of human languages. Several domainindependent efforts have been made for NLP, and this paper focuses on how computers can be able to understand semantic meaning of building regulations to intelligently automate rule interpretation process. This paper proposes a semantic analysis process of regulatory sentences and its utilization for rule checking system. The proposed process is composed of following steps: 1) learning semantics of words and sentences, 2) utilization of semantic analysis. For semantic analysis, we use word embedding technique which converts meaning of words in numerical values. By using those values, computers can extract related words and classify the topic of sentences. The results of the semantic analysis can elaborate the interpretation with domain-specific knowledge. This paper also shows a demonstration of the proposed

[1]  Jeffrey Dean,et al.  Efficient Estimation of Word Representations in Vector Space , 2013, ICLR.

[2]  Scharolta Katharina Siencnik.  Adapting word2vec to Named Entity Recognition , 2015, NODALIDA.

[3]  Yoon Kim,et al.  Convolutional Neural Networks for Sentence Classification , 2014, EMNLP.

[4]  Nora El-Gohary,et al.  Integrating semantic NLP and logic reasoning into a unified system for fully-automated code checking , 2017 .

[5]  Jeffrey Dean,et al.  Distributed Representations of Words and Phrases and their Compositionality , 2013, NIPS.

[6]  Ghang Lee,et al.  Requirements for computational rule checking of requests for proposals (RFPs) for building designs in South Korea , 2015, Adv. Eng. Informatics.

[7]  Charles M. Eastman,et al.  Automatic rule-based checking of building designs , 2009 .

[8]  Geoffrey E. Hinton,et al.  Learning representations by back-propagating errors , 1986, Nature.

[9]  Hyunsoo Lee,et al.  Translating building legislation into a computer-executable format for evaluating building permit requirements , 2016 .

[10]  George A. Miller,et al.  WordNet: A Lexical Database for English , 1995, HLT.

[11]  Eilif Hjelseth,et al.  Capturing normative constraints by use of the semantic mark-up RASE methodology , 2011 .

[12]  Charles M. Eastman,et al.  BIM Handbook: A Guide to Building Information Modeling for Owners, Managers, Designers, Engineers and Contractors , 2008 .

[13]  Quoc V. Le,et al.  Distributed Representations of Sentences and Documents , 2014, ICML.

[14]  Guigang Zhang,et al.  Deep Learning , 2016, Int. J. Semantic Comput..

[15]  Petr Sojka,et al.  Software Framework for Topic Modelling with Large Corpora , 2010 .

[16]  Eunjeong Lucy Park,et al.  KoNLPy: Korean natural language processing in Python , 2014 .

[17]  Charles M. Eastman,et al.  Classification of rules for automated BIM rule checking development , 2015 .