Construction Research Congress 2020 1127

A Comparative Analysis of Five Rule-Based Model Checking Platforms

Yong-Cheol Lee1; Pedram Ghannad2 ; Johannes Dimyadi3 ; Jin-Kook Lee4; Wawan Solihin5;
and Jiansong Zhang6
1
Assistant Professor, Bert S. Turner Dept. of Construction Management, Louisiana State Univ.,
Baton Rouge, LA. E-mail: [email protected]
2
Ph.D. Student, Bert S. Turner Dept. of Construction Management, Louisiana State Univ., Baton
Rouge, LA. E-mail: [email protected]
Downloaded from ascelibrary.org by University of Liverpool on 08/27/23. Copyright ASCE. For personal use only; all rights reserved.

3
Post-Doctoral Researcher, Dept. of Computer Science, Univ. of Auckland, New Zealand. E-
mail: [email protected]
4
Assistant Professor, Dept. of Interior Architecture, Yonsei Univ., Seoul, Republic of Korea
(corresponding author). E-mail: [email protected]
5
novaCITYNETS Pte. Ltd. Singapore, Singapore. E-mail: [email protected]
6
Assistant Professor, School of Construction Management Technology, Purdue Univ., West
Lafayette, IN. E-mail: [email protected]

ABSTRACT
With the boom in building information modeling, the need for an automated rule-based
design checking has increased to elevate the speed of design requirement evaluation pertaining to
building rules, regulations, and other necessary requirements. Over the past decade, various
efforts to develop an efficient rule-based checking process and platform have been made in
different countries. Even though there is one commercial tool named Solibri Model Checker,
none so far has been advanced enough to fully automate the design checking process and meet
the requirements of the industry. This study aims at investigating and comparing the current rule-
based design checking approaches including IfcDoc, KBim, ePlanCheck, ACABIM, and
SNACC, identifying the similarities and the differences of the existing approaches. The authors
who have been involved in each process of rule-checking tool development had a long discussion
regarding their experience and lessons learned and found that the five approaches were based on
a similar process and method to translate requirements, define rules, check design procedures,
develop validation features, and report checking results.

INTRODUCTION
The process of design compliance checking pertaining to regulatory standards, codes, and by-
laws is an essential part of the procedure for obtaining various permits required throughout the
entire lifecycle of a building project. The advent of technology into the building industry has led
various designs, complicated structures, and complex building projects that serve multiple
purposes. The standards, codes, and bylaws to be followed for designing and constructing a
building also have expanded and developed in order to incorporate explicit details of new project
environment or requirements to comply with the current trends in the AEC industries and
provide guidelines to support the complexity of these projects. However, the current manual
checking processes of design for compliance with standards, codes, and bylaws are not only
immensely time-consuming but challenging for project participants as well as the enforcers (Lee
et al., 2015). The rules and regulations are often complex and abstract, and thus, open to different
interpretations, leading to ambiguity (Khemlani, 2005) and overruns on cost and schedule. The
numerous requirements and regulatory components of building, facility, or infrastructure projects

© ASCE

Construction Research Congress 2020

 Construction Research Congress 2020 1128

always bring a huge burden to project participants, particularly ones such as an architect who are
involved in the early design phase (Ghannad et al, 2019). No matter how much time and effort
are invested in planning and checking whether these projects comply with all standards, codes,
and laws, yet an enormous number of requirements are frequently identified at the phase of
construction or later, resulting in significant additional cost and delay. In addition, abstract and
unclear regulatory statements require manual examinations that consequently cause different
interpretations and inconsistent applications. Particularly, heterogeneous information and
requirements of each building project frequently provoke controversy in rigorously applying
Downloaded from ascelibrary.org by University of Liverpool on 08/27/23. Copyright ASCE. For personal use only; all rights reserved.

rules in the project. Furthermore, since the design is subject to be iteratively exchanged and
shared with project participants such as an architect, a structural engineer, MEP designers,
facility managers, and others, it is demanding to iteratively evaluate if the design fulfills all
requirements and consistently maintain the quality of the design.
Despite these adverse situations, diverse domains have devoted to formalizing regulatory
information and establishing an automated rule-based checking process by bringing automation
into the compliance checking process, both for efficiency and accuracy. Previous projects have
distinct purposes and rule sources, which result in building a variety of rule checking features.
However, because of the nature of rule-based checking and the limited scope of implementable
rulesets, the authors have identified that most of the automated rule checking approaches entail
the similar processes and mechanism to translate and execute rules for validating design data.
This paper includes the experience-based descriptions and lessons learned from rule-based BIM
model checking processes of the authors. Each author has participated in the following five
prominent rule checking projects: IfcDoc-based MVD checking, KBIM, ePlanCheck, ACABIM
human-guided compliance audit system, and semantic natural language processing (NLP)-based
automated compliance checking (SNACC) system. Several research papers and reports regarding
these rule checking projects have been published, illustrating the targeted checking scopes and
objectives, detailed development processes, and their new intellectual contributions and practical
applications. By recognizing that these papers encompass numerous duplicates in processing
rules, addressing their manual translation, and building checking systems, the authors had several
discussions to share their project experiences and challenges and proposed diverse solutions to
handle the present issues. In this paper, with the lessons and knowledge from these rule checking
research studies and projects, the authors explore the mandatory processes, the identified
challenges, and the promising rule checking directions.

LITERATURE REVIEW
There has been a significant development in automated rule checking in the last decade. As a
representation of building designs and plans started converting from simple 2D drawings into a
3D model, automated rule-based checking developed into a new field of research with the BIM
technology (Solihin et al., 2017a). The major challenges in the area have always been the
conversion of rules and regulations written on paper into a format that is both understood by end-
users as well as computers. Also, the fact that each country has its own set of codes and
normative standards, with some countries also having regional and local codes and by-laws that
may override the general rules of the country poses a greater challenge. In addition, there is also
the ever-changing rules and regulations in the form of acts and amendments that need to be
constantly updated in order to be effective for rule checking systems.
One of the earliest attempts made at automated rule checking was the development of a
software application named BP-Expert. It was one of the earlier approaches made in the

© ASCE

Construction Research Congress 2020

 Construction Research Congress 2020 1129

CORENET Project for rule-checking in Singapore and was based on the use of Artificial
Intelligence and Feature-based Cad-technologies (Dimyadi and Amor., 2013). The CORENET
project aimed at creating an IT infrastructure that allowed the total integration of the four basic
processes of the building life cycle namely Design, Procure, Build and Maintain (Khemlani,
2005). The attempt was unsuccessful because of various reasons including the use of 2D
drawings instead of a 3D model, non-flexibility to handle errors in data, limited ability and use as
it could not be applied to all code clauses, disputes regarding the ownership of the applications
and the cumulative efficiency and accuracy of the application (Khemlani, 2005). Other attempts
Downloaded from ascelibrary.org by University of Liverpool on 08/27/23. Copyright ASCE. For personal use only; all rights reserved.

at rule checking included the SASE system (Standards Analysis, Synthesis, and Expression) and
REGNET. The aim of SASE was to provide a comprehensive structure for families of related
codes while REGNET focused on determining the applicability of different codes for a given
building condition, based on a question-answer user interface (Fenves, 1987). Apart from
REGNET, there were also SICAD and SPEX that used SASE as their rule base. Though these
efforts helped to lay the foundation for the logical structure of codes from a rule-based and
organizational perspective, they failed to address the automated application of rules to a digital
building representation to a satisfactory level (Eastman et al., 2009).
During this early research era in the field of building rule checking, several platforms have
developed to support the implementation aspect of rule checking systems. They all complied to
the rules of the industry foundation classes (IFC) building model data. One of the commercial
applications is the Solibri Model Checker, which is broadly utilized in many countries. It is a
Java-based platform that reads the IFC model and maps it to an internal structure, thereby aiding
in the access and processing of rule checking. It includes several built-in functions for the pre-
checking model in terms of shape, overlaps, name and attribute conventions, object existence,
etc. (Greenwood et al., 2010). Other built-in functions help in automatic viewing of checking
issues, checking accessibility, check for various spaces in the building and fire code exit path
distance check (Eastman et al., 2009). DesignCheck developed in Australia, and SMARTCodes
in the US were two other example approaches from early 2000 that have not survived the test of
time (Dimyadi and Amor., 2013).

METHODOLOGY
In order to achieve the goal of this research, the authors provided their experience and
development processes to identify the scope, purpose, motivation, and processes of the five rule-
checking methods. The authors of this paper, who has been closely involved in these five
research projects, had several discussions to share their project experiences and challenges and
proposed diverse solutions to handle the present issues. By recognizing the similarities of the
methods, the authors conducted a comparative analysis focusing on rule translation and
definition processes, rule categorization processes, and rule execution processes.

OVERVIEW OF FIVE RULE-BASED BIM DATA CHECKING RESEARCH STUDIES

IfcDoc
IFC (Industry Foundation Classes) is a platform-neutral, open object-based file format
specification with a data model that is developed by buildingSMART, formerly the International
Alliance for Interoperability (IAI). With the advent of automation in rule checking, it was found
that data structure and information contained within the IFC model is not compatible with the
rule checking environment. Thus, a conversion phase is required for converting the IFC model

© ASCE

Construction Research Congress 2020

 Construction Research Congress 2020 1130

into a model containing information necessary for the rule checking platform (Pauwels et al.,
2011). An IFC View Definition or Model View Definition (MVD) is a subset of IFC schema that
is created and published in the neutral MvdXML format and is required to satisfy Exchange
Requirements (ER) of specific applications in the AEC industry (Lee et al., 2019a; 2019b). The
methods used to define such ER are defined in the ISO 29481 Information Delivery Manual
(IDM) (BuildingSMART). An IDM defines the scope and working processes of the models in
order to document the possible information needs and constraints as exchange requirements and
business rules. The plain text is then structured according to exchange requirement models and
Downloaded from ascelibrary.org by University of Liverpool on 08/27/23. Copyright ASCE. For personal use only; all rights reserved.

linked with specific IFC entities, attributes, and types as MVDs (Lee et al., 2019a). The major
motivation behind the development of the IfcDoc tool with MvdXML specification was to create
an open standard for MVD and IFC validation that could provide a low-cost rule checking
environment. The MvdXML is an XML serialization of the IFC view definition that can be
generated by the official IfcDoc tool given an ER specification (Lee et al., 2018). Any XML
editor can also be used to edit the file.

KBIM
KBim is a Korean government initiative to enhance the productivity, performance, and
overall quality of building design by using BIM, and some KBim-series applications have been
developed and are still ongoing. KBimCode is one of KBim-series apps as a project to represent
the Korean Building Act in an explicit and computable form (Park & Lee, 2016). This platform
was specifically developed for automated rule checking of the buildings constructed in Korea, as
well as a part of the building permit system of Korea (Seumter, 2019). The Korea Building Acts
are the most important and primary regulations established by the government that needs to be
followed while constructing any building in Korea (Park et al., 2016). It includes a vast range of
topics related to the building such as design, structure, material, fire safety, facility management,
etc. A logic rule-based approach was developed for translating and analyzing the sentences
written in the building code and into a computer-readable language like Java or Python. It
incorporated a visual language-based approach, utilizing visual symbols for translating rules,
known as KBim Visual Language (KBVL). It identifies the features of building regulations,
visualizes the components in the sentence as per their functions and grammar characteristics and
demonstrates them through the visual language.

ePlanCheck
ePlanCheck is one of the earliest complete implementations of rule checking systems that
was initiated by the government of Singapore in 1995. It was developed as a part of the
CORENET project by the Singapore Ministry of National Development and was commissioned
by the Singapore Building and Construction Authority (BCA). The CORENET project was
divided into three modules in the Design phase (Eastman et al., 2009): e-Submission, e-Info, and
e-PlanCheck. e-PlanCheck currently deals with building rules related to building services,
barrier-free access, fire code, environmental health, household, public housing and parking.
Building services include electrical, fire alarm system, fire sprinkler system, raising main and
fire hydrant system, ventilation, sanitary, plumbing, and drainage system, surface water drainage,
gas pipe system and water services. ePlanCheck is one of the most developed rule-checking
method currently available. The main aim of the CORENET project was to allow rapid and
seamless sharing of information between different parties on a building project, thus saving time
and cost, and improving quality. The main motivation behind developing e-PlanCheck was to

© ASCE

Construction Research Congress 2020

 Construction Research Congress 2020 1131

bring in automation into the building code-checking process and discard the traditional manual
system of paper-based code-checking.

Table 1. Comparison Regarding Rule Translation and Definition Processes

Systems Rule Translation and Definition Processes
IfcDoc - Rules are defined as Concept Template and Concept Levels.
- Concepts have a tree structure.
- Tree structure denotes the entities and attributes.
Downloaded from ascelibrary.org by University of Liverpool on 08/27/23. Copyright ASCE. For personal use only; all rights reserved.

ePlanCheck - Use semantic objects in the FORNAX,

- FORNAX objects extend IFC objects to include data
functions and algorithm relevant to the domain of building
codes.
KBIM - Rules are restructured and represented as KBimCode Using
KBimLogic.
- A visual language, KBVL, is also deployed in KBIM to
automatically generate a computer-readable code using
KBimLogic.
ACABIM -The semantics of legal requirements are translated semi-
automatically into computable rules using a systematic
approach.
- Rules are represented using the open XML data structure,
specifically LegalRuleML which is built on top of RuleML
and in conjunction with the companion open standard
LegalDocML to represent the literal content of each legal
document.
SNACC System - Semantic NLP algorithms automatically extract regulatory
information from building codes.
- Pattern matching rule-based algorithms using syntactic and
semantic properties of the extracted regulatory information
automatically transform extracted regulatory information into
logic rules.

ACABIM
ACABIM is a recent commercialization of automated compliance audit research, being
undertaken by Compliance Audit Systems Limited in New Zealand, partnering with the
commercial wing of the University of Auckland. The underlying philosophy of ACABIM is to
automate current manual compliance audit processes focusing on trivial tasks that are tedious for
human experts to perform, thereby allowing human experts to have more time to assess
performance-based design that requires human expertise and creativity. The research behind
ACABIM has been published in journals and conference proceedings (Dimyadi and Amor.,
2017). The ACABIM framework supports three main input components, namely the Open BIM
IFC, an open standard workflow model representing the compliance audit procedures (CAP) or
the logical compliance pathways, and emerging open standard LegalRuleML representing
normative rules (LKM). Additionally, it also supports supplementary human input as well as
interfacing with simulation tools. The initial implementation incorporates the translation of a
core set of compliance documents from the New Zealand Building Code including provisions

© ASCE

Construction Research Congress 2020

 Construction Research Congress 2020 1132

related to fire safety, ventilation, sanitary facilities, and energy efficiency, to serve the priority
consenting requirements.

The Semantic Natural Language Processing (NLP)-based Automated Compliance

Checking (SNACC) System
The semantic natural language processing (NLP)-based automated compliance checking
(SNACC) system is the first system that not only automates the compliance reasoning process
that checks a building design with building code requirements but also automates the extraction
Downloaded from ascelibrary.org by University of Liverpool on 08/27/23. Copyright ASCE. For personal use only; all rights reserved.

and encoding of regulatory information from textual documents into a computable representation
(Zhang and El-Gohary, 2013; 2017). The system is built upon NLP, EXPRESS data processing,
and logic reasoning technologies. A semantic, logic-based representation was developed that is
generalized and flexible. Specifically, logic rules are used to represent regulatory information
and logic facts are used to represent design information. Automated regulatory information
extraction and transformation algorithms were developed based on deep NLP that integrated
automated compliance checking (ACC)-specific knowledge, architectural, engineering, and
construction (AEC) domain knowledge, and linguistic knowledge. For processing building
design information, EXPRESS data processing techniques are used for automatically extracting
design information from BIMs and transforming the extracted design information into logic facts
(Zhang and El-Gohary, 2015).

COMPARISON OF FIVE RULE-BASED BIM DATA CHECKING APPROACHES

This study analyzed the existing tools according to the following three processes: (1) rule
translation and definition processes, (2) rule categorization, and (3) rule execution. A
comparative analysis has been carried out based on these four stages. The summarized results
have been provided as Table 1-3.

LESSONS LEARNED AND CHALLENGES

Based on the comparative analysis of the five methods, several important lessons were
learned about the rule-checking processes and limitations.
 The similarity of the development phases: This study was designed to gain multiple
perspectives on how different research projects regarding rule-based BIM model
checking conducted the rule translation, categorization, and execution and addressed the
challenges of the rule-checking process. It is evident that there is a uniformity in the
process of the framework development but using different methods for each process: (1)
rule translation and definition processes, (2) rule categorization, and (3) rule execution.
This similarity between various methods validates that these steps are required to develop
a rule-based BIM design framework.
 BIM-based application: Although all the methods have their own procedure and
specifications, utilizing BIM is an inseparable part of all studies. In particular, as an open
standard, IFC plays a pivotal role in BIM model checking.
 Lack of flexibility: One of the major shortcomings of current rule-checking platforms is
the lack of flexibility. Using hard-coded rules and “Black box” for rule-checking in
several commercial platforms lead to lack of transparency and flexibility in these
methods. The five methods studied in this research tried to overcome this gap by using
open standards, software-independent process, visual language programming, and natural

© ASCE

Construction Research Congress 2020

 Construction Research Congress 2020 1133

language processing, but its inherent problem still remains.

 Limitation in rule translation: Even though the five studies proposed a way to
automatically translate written regulations into implementable formats using open
standards such as XML and LegalRuleML, and semantic-based method such as NLP, it
still requires significant improvement for covering a wide range of rules including
abstract definitions and various project requirements regardless of rule types, origins, and
domains. In addition, the fact that each country has its own set of codes and normative
standards, with some countries also having regional and local codes and by-laws that may
Downloaded from ascelibrary.org by University of Liverpool on 08/27/23. Copyright ASCE. For personal use only; all rights reserved.

override the general rules of the country poses a greater challenge. In addition, there is
also the ever-changing rules and regulations in the form of acts and amendments and the
utilized methodology need to be capable of being constantly updated in order to be
effective for rule checking systems.

Table 2. Comparison Regarding Rule Categorization

Systems Rule Categorization
IfcDoc - Checking data for the existence and cardinality
- Checking the content of values
- Uniqueness of values
- Checking the if-then conditional dependency and consistency
among them
ePlanCheck - Check the current IFC model information
- Check the property sets of the IFC extensions
- Check the information that is derived from the IFC model
using FORNAX objects.
KBIM - Arithmetic Logic Units (ALUs) (Single or multiple)
- Condition Statement (IFTHEN-ELSEIF-ELSE logic)
- KBimCode Object Model (KOM) (user-defined virtual objects
with desired rules as constraints)
ACABIM
- Using an open standard workflow model with domain-specific
query language to represent the logical compliance pathways
and audit procedures.
- The workflow model can extract information from the IFC
model and compliance rules from either the official central
repository for national norms or an independently served
repository for contractual requirements.
SNACC System - Semantic NLP algorithms automatically extract regulatory
information from building codes
- Text classification was used to classify provisions of building
codes into different types such as quantitative requirements and
existential requirements
- Converted logic rules were divided into primary logic clauses
and secondary logic clauses depending on the level of
importance of the information they carry

© ASCE

Construction Research Congress 2020

 Construction Research Congress 2020 1134

Table 3. Comparison Regarding Rule Execution

Rule-checking
Rule Execution
systems
IfcDoc - The EXPRESS schema of IFC is transferred to an Eclipse
Modelling Framework (EMF)
- Extracts related IFC objects and their features by their
names that are defined in the mvdXML file.
- Evaluating the values and checking
Downloaded from ascelibrary.org by University of Liverpool on 08/27/23. Copyright ASCE. For personal use only; all rights reserved.

ePlanCheck FORNAX™ ePlanCheck uses its own checking engine that

integrates logic and geometric reasoning for the rule
execution.
KBIM - KBimCode is software-independent and applicable in
various BIM assessment tools.
- For example, one of the developed BIM assessment tools
named KBimAssess-lite can be used.
ACABIM -Using a dedicated process engine to audit the data collected
by the workflow model for compliance
- Using an internal mapping process in conjunction with a
data dictionary to match objects within the IFC model with
those defined in the rules or Legal Knowledge Model (LKM)
SNACC System - Rule execution is enabled by the logic reasoner of logic
programming platforms
- Logic facts that represent design information are iterated
through to find matching with logic rules that encode
regulatory requirements, matched logic facts are used to
instantiate the corresponding logic rules

CONCLUSION
As BIM technology has been widely used in our industry, automating BIM design checking
and evaluation is inevitable. Since each project has been set out to pursue different purpose with
different application scope, its approach to rule validation, evaluation, and execution have
developed differently, adopting appropriate methods as necessary. However, it is evident that
each rule-checking implementation also entails homogeneous steps as well as encounters similar
challenges. It may say that the identified processes are required for building an automated rule-
based BIM checking framework. In addition, the challenges come from inherent issues of the
building design industry and regulatory compliance checking with BIM models. In particular, the
conversion of rules from human-readable format to computer-readable format has been
recognized as the main challenges of these methodologies. Using open standards and semantic-
based methods like NLP were two dominant approaches used by the five research projects.
However, the authors believe that the challenges will be overcome soon by advanced technology
or methodologies for boosting building design automation and rapid compliance checking.

REFERENCES
BuildingSMART. (2014). The CORONET project in Singapore. BuildingSMART Case Studies,
BuildingSMART.

© ASCE

Construction Research Congress 2020

 Construction Research Congress 2020 1135

BuildingSMART. (2019). http://www.buildingsmart-tech.org/specifications/ifc-view-definition

Khemlani, L. (2004). Analysis, research, and review of AEC technology. Retrieved September
25, 2010.
Dimyadi, J., & Amor, R. (2013). Automated building code compliance checking–where is it at.
Proceedings of CIB WBC, 6.
Dimyadi, J., & Amor, R. (2017). Automating Conventional Compliance Audit Processes. In IFIP
International Conference on Product Lifecycle Management (pp. 324-334). Springer, Cham.
Eastman, C., Lee, J. M., Jeong, Y. S., & Lee, J. K. (2009). Automatic rule-based checking of
Downloaded from ascelibrary.org by University of Liverpool on 08/27/23. Copyright ASCE. For personal use only; all rights reserved.

building designs. Automation in construction, 18(8), 1011-1033.

Fenves, S.J., Wright, R.N., Stahl, F.I., and Reed, K.A. (1987). “Introduction to SASE: standards
analysis, synthesis and expression.” U.S. Dept. of Commerce, Washington, DC.
Ghannad, P., Lee, Y. C., Dimyadi, J., & Solihin, W. (2019). Automated BIM data validation
integrating open-standard schema with visual programming language. Advanced Engineering
Informatics, 40, 14-28.
Greenwood, D., Lockley, S., Malsane, S., & Matthews, J. (2010). Automated compliance
checking using building information models. In the Construction, Building and Real Estate
Research Conference of the Royal Institution of Chartered Surveyors, Paris 2nd-3rd
September. RICS.
Lee, Y. C., Eastman, C. M., & Lee, J. K. (2015). Automated rule-based checking for the
validation of accessibility and visibility of a building information model. In Computing in
Civil Engineering 2015 (pp. 572-579).
Lee, Y. C., Eastman, C. M., & Solihin, W. (2018). Logic for ensuring the data exchange integrity
of building information models. Automation in Construction, 85, 249-262.
Lee, Y. C., Ghannad, P., & Lee, J. K. (2019a). Reusability and Its Limitations of the Modules of
Existing BIM Data Exchange Requirements for New MVDs. In Advances in Informatics and
Computing in Civil and Construction Engineering (pp. 103-110). Springer, Cham.
Lee, Y. C., Solihin, W., & Eastman, C. M. (2019b). The Mechanism and Challenges of
Validating a Building Information Model regarding data exchange standards. Automation in
Construction, 100, 118-128.
Method manual, Open BIM based Technological Environment for Building Design Quality
Enhancement, On-line: http://designitlab.kr/bim/kbim/method/classification.asp
Park, S., & Lee, J. K. (2016). KBimCode-based applications for the representation, definition
and evaluation of building permit rules. In ISARC. Proceedings of the International
Symposium on Automation and Robotics in Construction (Vol. 33, p. 1). Vilnius Gediminas
Technical University, Department of Construction Economics & Property.
Park, S., Lee, Y. C., & Lee, J. K. (2016). Definition of a domain-specific language for Korean
building act sentences as an explicit computable form. Electronic Journal of Information
Technology in Construction, 21(2016).
Pauwels, P., Van Deursen, D., Verstraeten, R., De Roo, J., De Meyer, R., Van de Walle, R., &
Van Campenhout, J. (2011). A semantic rule checking environment for building performance
checking. Automation in Construction, 20(5), 506-518.
Seumter. (2019)., https://www.eais.go.kr, Accessed June 2019.
Solihin, W., Dimyadi, J., Lee, Y. C., Eastman, C., & Amor, R. (2017a). The critical role of the
accessible data for BIM based automated rule checking system. In Proceedings of the Joint
Conference on Computing in Construction (JC3) (Vol. 1, pp. 53-60).
Solihin, W., Eastman, C., & Lee, Y. C. (2017b). Multiple representation approach to achieve

© ASCE

Construction Research Congress 2020

 Construction Research Congress 2020 1136

high-performance spatial queries of 3D BIM data using a relational database. Automation in

Construction, 81, 369-388.
Zhang, J., and El-Gohary, N. (2013). "Semantic NLP-based information extraction from
construction regulatory documents for automated compliance checking." J. Comput. in Civ.
Eng., 30(2), 04015014.
Zhang, J., and El-Gohary, N. (2015). "Automated extraction of information from building
information models into a semantic logic-based representation." Proc., 2015 ASCE Intl.
Workshop on Comput. in Civ. Eng., ASCE, Reston, VA, 173-180.
Downloaded from ascelibrary.org by University of Liverpool on 08/27/23. Copyright ASCE. For personal use only; all rights reserved.

Zhang, J., and El-Gohary, N. (2017). "Integrating semantic NLP and logic reasoning into a
unified system for fully-automated code checking." Automation in Construction, 73, 45-57.

© ASCE

Construction Research Congress 2020