Content uploaded by Isuru Eranga
Author content
All content in this area was uploaded by Isuru Eranga on Dec 06, 2024
Content may be subject to copyright.
Saparamadu, P.V.I.N., Jayasena, H.S. and Eranga, B.A.I., 2024. Blueprint for a natural language
processing powered nexus for regulatory and legal landscape in construction. In: Sandanayake, Y.G.,
Waidyasekara, K.G.A.S., Ranadewa, K.A.T.O. and Chandanie, H. (eds). Proceedings of the 12th World
Construction Symposium, 9-10 August 2024, Sri Lanka. pp. 306-317. DOI:
https://doi.org/10.31705/WCS.2024.24. Available from: https://ciobwcs.com/papers/
BLUEPRINT FOR A NATURAL LANGUAGE
PROCESSING POWERED NEXUS FOR
REGULATORY AND LEGAL LANDSCAPE IN
CONSTRUCTION
P.V.I.N. Saparamadu
1
, H.S. Jayasena
2
, and B.A.I. Eranga
3
ABSTRACT
The recent exponential advancements in Natural Language Processing (NLP) are
catalysing a paradigm shift in the world, directing the construction industry towards an
era of smart construction. The proficiency of NLP in comprehending and assimilating
vast quantities of human language data aligns aptly with the construction sector’s
exigency for enhanced management of its unstructured textual data. Given the frequent
alterations in regulatory frameworks and the dispersed nature of project data, there
arises a compelling need for a Natural Language Processing Powered Compliance
Management Nexus (NLP-PCMN), which facilitates expedited access to consolidated
information via mobile platforms. This study aims to develop a blueprint for
implementing an NLP-PCMN in the construction industry. By conducting semi-
structured interviews with 20 experts spanning the domains of construction and Artificial
Intelligence (AI) alongside a focus group to outline the technological framework of the
NLP-PCMN, the research underscores the need to implement such a system. The
envisaged system is poised to address challenges such as navigating contract clauses,
correspondence analysis and ensuring legal compliance with planning and building
codes and legal provisions. The proposed NLP-PCMN presents a comprehensive
solution integrating these features through large language models that work as a
question-and-answering system. Key findings include the necessity of automating the
regulatory and legal data in construction, stakeholder empowerment through NLP-
PCMN, identifying the nodes of the NLP-PCMN and the technical blueprint to implement
the NLP-PCMN.
Keywords: Artificial Intelligence (AI); Construction Law; Natural Language
Processing (NLP); Smart Construction.
1. INTRODUCTION
As compelling evidence of the contemporary paradigm shift in AI technologies, the
global AI in construction market size was valued at USD 696 million in 2023 and is
projected to reach fivefold by 2032 (360iResearch, 2023). Among those technologies,
Natural Language Processing (NLP) technology made a significant leap in the last five
1
Research Student, Department of Building Economics, University of Moratuwa, Sri Lanka,
ishini@concolabs.com
2
Senior Lecturer, Department of Building Economics, University of Moratuwa, Sri Lanka,
suranga@uom.lk
3
Lecturer, Department of Building Economics, University of Moratuwa, Sri Lanka, isurue@uom.lk
Blueprint for a natural language processing powered nexus for regulatory and legal landscape in
construction
Proceedings The 12th World Construction Symposium | August 2024 307
years as AI exceeded human performance levels on basic reading comprehension
benchmarks (Stone et al., 2021). NLP is a subfield of AI for difficult language-related
problems, such as machine translation, question-answering and summarisation (Lauriola
et al., 2022). The recent trend of NLP applications is a consequence of introducing Large
Language Models (LLMs) such as OpenAI’s ChatGPT, Google’s BERT and MT5 (Zhang
et al., 2022). LLMs are a subset of NLP, representing advanced models that emerged in
2018 (Cambria & White, 2014).
Although NLP applications in various fields began in the 1960s, their adaptation to the
construction industry commenced in the 1990s (Khurana et al., 2023). The first
application of NLP to the construction industry was developed in 1989, aiming to aid
construction managers in retrieving vital information for decision-making (Khurana et
al., 2023). NLP applications have been developed for its thematic areas in the construction
industry. These include enhancing information and document management processes,
improving compliance management, evaluating public perceptions of construction
projects and optimising contract management (Hassan & Le, 2020).
Babatunde et al. (2023) revealed that proficient compliance management can be achieved
through NLP by enhancing contract accuracy, efficiency and transparency. Legal and
regulatory compliance in the construction industry is of paramount importance due to the
complex and stringent laws governing this sector (Beach et al., 2015). Regulatory
compliance ensures that construction projects adhere to established standards, codes and
regulations. According to Marzouk and Enaba (2019), legal compliance includes contract
management, appropriate documentation and correspondence procedures. Automating
compliance processes in construction can streamline operations, reduce human error, and
enhance efficiency (Beach et al., 2015). Furthermore, an automated compliance
management system provides real-time monitoring and reporting capabilities (Beach et
al., 2020). Integrating automation into legal and regulatory compliance processes in
construction is essential for ensuring project success.
Parikh et al. (2023) documented the landmark court case that used NLP to form a judicial
decision for the first time. It is a result of the evolution of NLP technology in recent years,
driven by advancements in deep learning and machine learning (Khurana et al., 2023).
One of the key breakthroughs has been the development of powerful language models,
which have been trained on vast amounts of textual data that can capture the nuanced
relationships between words and the context in which they are used (Zhang et al., 2022).
Yan et al. (2020) emphasised that the legal and regulatory data in the construction domain
are available as unstructured data sources such as text documents. Therefore, efficient
and intelligent extraction and interpretation of this textual data is vital for the cost-
effective management of projects (Wu et al., 2022). NLP provides the solution for it
through the analysis of text structures and words (Nadkarni et al., 2011). Thus, with NLP
advancements continuing to unfold, it has become increasingly important for construction
professionals to rethink how to leverage it to enhance their practices and processes.
It was found that numerous models have been developed for legal and contractual
domains, utilising various advancements in NLP technology. For example, Beach et al.
(2020) and Lee et al. (2023) have contributed models geared towards regulatory
compliance automation, while Hand et al. (2021) and Lee et al. (2020) have focused on
models for contract management. After testing these models, the above-mentioned
literature found a significant increase in the accuracy of compliance management.
P.V.I.N. Saparamadu, H.S. Jayasena, and B.A.I. Eranga
Proceedings The 12th World Construction Symposium | August 2024 308
Although several NLP models have been developed in the literature, NLP adoption in the
construction industry is still in its infant stage (Madan & Ashok, 2023). Wu et al. (2022)
identified that the lack of awareness in the industry on utilising LLMs for their business
purposes was a fundamental reason for the current state of adoption. Adding to the
statement above, Madan and Ashok (2023) highlighted the limited adoption of NLP
because there is not a widely recognised NLP model in commercial use, and the existing
models aren’t feasible for practical application in the construction sector.
Hence, it can be suggested that an ideal approach to harnessing the potential of NLP
entails establishing a robust nexus comprising specialised tools tailored to be used by all
stakeholders of the industry. This nexus-oriented strategy facilitates seamless integration
and synergy among various NLP capabilities, avoiding the need for disparate toolsets
(Mitchell & Mancoridis, 2006). A single platform housing all software tools is crucial for
several reasons. It simplifies the management process by reducing the complexity and
overhead associated with managing multiple systems. This approach also reduces the risk
of data inconsistencies by storing and processing data within a single system (Mitchell &
Mancoridis, 2006). Thus, there is a timely need for an NLP-powered nexus for effective
adoption in the construction industry.
While ChatGPT has demonstrated significant capabilities in NLP, it is insufficient as a
standalone solution for specialised applications such as legal research. ChatGPT’s
general-purpose design lacks the domain-specific knowledge (Parikh et al., 2023).
Studies have shown that domain-specific NLP models significantly outperform general
models in specialised fields (Jurafsky & Martin, 2021). Therefore, a tailored approach
incorporating domain-specific NLP tools is essential.
Despite the growing interest in NLP tools, there remains a shortage of research on their
application across various domains, including the construction industry. Jallan et al.
(2019) studied the development of an NLP model to conduct a comprehensive survey of
legal cases; however, they used statistical algorithms rather than LLMs. However, a study
by Moon et al. (2022) used recent developments in NLP to review specifications.
Furthermore, while recent research by Shaikh and Gohar (2024) has investigated the use
of chatbots in contract management, it does not offer a comprehensive solution for the
entire legal landscape. Therefore, this study seeks to pioneer such a concept of an
accessible legal database tailored explicitly for the construction industry. Hence, it aims
to develop a blueprint for implementing an NLP-powered Construction Management
Network (NLP-PCMN). The paper begins with a literature review and continues by
investigating the need for an NLP-PCMN. Finally, it presents an implementation
blueprint detailing the architecture and key components necessary for its development.
2. LITERATURE REVIEW
2.1 NLP APPLICATIONS IN THE CONSTRUCTION INDUSTRY
Studies suggest that NLP has been used in the construction industry for following
application scenarios, including filtering information, organising documents, using expert
systems and automating compliance checking (Wu et al., 2022). From the literature
analysis, 91 NLP models were identified for various domains, as illustrated in Figure 1.
The models discussed in the literature were developed using obsolete advancements in
NLP, such as rule-based techniques, probability models and neural networks. Among the
main application areas, NLP for contract management accounted for 31%, making it the
Blueprint for a natural language processing powered nexus for regulatory and legal landscape in
construction
Proceedings The 12th World Construction Symposium | August 2024 309
domain with the highest number of applications. Compliance checking application
domain constituted 23% of the NLP models. Thus, the NLP models developed for the
regulatory and legal landscape of the construction industry amounted to 54% of all
models developed. Figure 1 shows the distribution of these models across the literature.
Figure 1: Distribution of NLP models across application areas (Source: Developed by authors)
2.2 STATE OF THE ART OF NLP MODELS IN CONSTRUCTION
Recent advancements in NLP have revolutionised the construction industry’s approach
to contract analysis and management. Notably, Padhy et al. (2021) demonstrated an 80%
increase in efficiency with an NLP model designed to detect exculpatory sentences. As
validated by Hand et al. (2021), F1 scores exceeding 70% indicate a reliable and effective
model. Lee et al. (2019) and Lee et al. (2020) further demonstrated the competence of
NLP in automatically detecting problematic clauses with impressive F1 scores of 81.8%
and 80%, respectively. These findings highlight the robustness of NLP in scrutinising
contractual documents, flagging critical clauses, and enhancing decision-making
processes in the construction industry.
2.3 REGULATORY AND LEGAL CONSIDERATIONS IN CONSTRUCTION
To formulate a nexus facilitating efficient regulatory and legal compliance within
construction projects, it is crucial to identify the textual data necessitating analysis.
Through literature analysis, various types of data were identified, as outlined in Table 1.
P.V.I.N. Saparamadu, H.S. Jayasena, and B.A.I. Eranga
Proceedings The 12th World Construction Symposium | August 2024 310
Table 1: Sources of textual data in a construction project
Citation
Sources of Textual Data
Soibelman et al.,
2008
conditions of contracts, specifications, change orders, requests for
information, meeting minutes
Kelley, 2012
Statutes and ordinances, agency regulations, international treaties, case
law, contract clauses
Szewc, 2022
environmental protection law, civil law regulations, public procurement
regulations, property law, planning law
Murdoch &
Hughes, 2002
insurance law, contract law, dispute resolution procedures, case law,
standard conditions, contract data, construction service agreements,
procurement law
Marzouk & Enaba,
2019
contract, variation order log, site instruction, progress reports, request for
information log, cost schedule data, claim data
3. RESEARCH METHODOLOGY
Qualitative research, particularly grounded in interpretivism, stands as the optimal
approach for the development of an NLP-PCMN. At the core of interpretivism lies the
belief that reality is subjective, emphasising the importance of understanding the unique
perspectives of the individuals involved in the data collection (Potrac et al., 2014). In the
context of NLP-PCMN, this study aims to find a speculative ideal for a prospective
technology that is well suited to the interpretive philosophy. By embracing qualitative
methodologies, the fluidity of different experiences can be seen while shedding light on
emergent patterns (Saunders et al., 2018).
The research process encompassed two phases of data collection. Initially, 20 semi-
structured interviews were conducted to ascertain the need to implement a nexus for
construction management within the construction domain. The interview questions were
designed to understand the necessity of automating activities suitable for NLP integration
and identify the nodes of the nexus. These experts were purposively selected to ensure a
balanced representation across construction, digitalisation and AI industries. Table 2
provides an overview of the experts’ profiles in the study.
Table 2: Profile of the experts
Nr
Designation/ Field
Experience
(Years)
Country
R1
Director - Consultant QS
22
Sri Lanka
R2
Director – Contractor QS
18
Sri Lanka
R3
Engineer - Transportation Sector
18
Australia
R4
LLM Developer for Procurement
10
Australia
R5
NLP model developer for Translations
8
USA
R6
Consultant QS/5D BIM Agent
32
Sri Lanka
R7
Professor – Construction Law
20
UK
R8
Construction Contract Manager
24
UK
R9
Construction Data Analyst
8
UK
R10
Construction Project Manager
18
Sri Lanka
R11
Expert Witness in Delay Analysis
32
Sri Lanka
Blueprint for a natural language processing powered nexus for regulatory and legal landscape in
construction
Proceedings The 12th World Construction Symposium | August 2024 311
Nr
Designation/ Field
Experience
(Years)
Country
R12
Development Manager
15
UK
R13
Contract Manager
10
Nigeria
R14
Construction Lawyer
9
UK
R15
Construction Automation Professor
20+
Australia
R16
Data Science Professor
30+
Germany
R17
Commercial Manager
15+
UK
R18
Claims Specialist
27
UAE
R19
Contract Specialist
24
UAE
R20
Chief Executive Officer/ Architect
12
UK
Following the analysis of the requirements outlined by these experts, the need to establish
an NLP-PCMN was substantiated. Subsequently, a focus group was convened involving
experts R4, R5, R9, and R16, proficient in AI model development, to formulate the
blueprint of an NLP-PCMN. Qualitative data obtained from the discussions were
subjected to content analysis, with the software NVivo 12 software facilitating systematic
analysis and interpretation of the data.
4. RESEARCH FINDINGS AND ANALYSIS
4.1 IMPERATIVE FOR IMPLEMENTING AN NLP-PCMN
The importance of implementing an NLP-PCMN was investigated through expert
interviews. A consensus among all respondents emphasised the criticality of effective
construction management, necessitating the storage and management of a vast repository
of indexed textual data within the construction domain. R1 interpreted this by stating that
“a human can’t store all the information and access the information all together.” R2,
R8, R10 and R13 confirmed this by stating that this complexity can overwhelm humans,
especially when sources precede the other. However, as explained by R5, AI is capable
of “Long Short-Term Memory that can access a large amount of data for a short period.”
Supporting this notion, nearly all respondents suggested that automation addresses this
challenge by alleviating reliance on human memory, thereby ensuring that critical points
are not overlooked. However, R14 presented a contrasting viewpoint, arguing that human
interaction remains indispensable as not all information may be pertinent to resolving a
legal matter. This contention was contested by R4, who proposed a model capable of
training on identifying precedence.
Moreover, R7, R8, R9, R17, and R20 highlighted that many professionals engaged in
contract management lack proficiency in legal affairs, often necessitating reliance on
legal experts for guidance. This dependency can result in delays, costs and inefficiencies.
Nevertheless, integrating legal considerations into automation fosters self-empowerment
among professionals and minimises inefficiencies. Furthermore, R11 and R14
emphasised the significance of staying current with case laws, albeit interpreting them
amidst evolving conditions and constantly changing planning regulations that can pose
challenges. Additionally, adherence to various building codes, which may vary by
location, is vital. As suggested by R20, automation streamlines this process by
centralising and organising information, thereby enhancing accessibility.
P.V.I.N. Saparamadu, H.S. Jayasena, and B.A.I. Eranga
Proceedings The 12th World Construction Symposium | August 2024 312
4.2 STAKEHOLDER EMPOWERMENT THROUGH NLP-PCMN
The expert interviews identified all types of users that will benefit from a proposed NLP-
PCMN and particular use cases of the nexus. Project managers at the site were referenced
most by experts for the professionals who would be empowered the most. Furthermore,
using this nexus as a legal handbook was predicted to be a use case for all its potential
users. R1 described it: “Although I might use books normally, I will use this in meetings
because turning pages is not nice.” Furthermore, R1, R12, R13, R15 and R17 addressed
that the chatbot will provide a significant quality of life for entry-level professionals in
the industry. Furthermore, 80% of the experts mentioned that clients will benefit from
this innovation as they can “stay in touch and understand the legal considerations of their
project.”
4.3 NODES OF THE NATURAL LANGUAGE-POWERED CONTRACT
MANAGEMENT NEXUS
An NLP-PCMN ideally should provide a comprehensive solution to tackle diverse legal
and regulatory compliance challenges within the construction industry. Through
collaboration with experts, the focus group discerned that the main features are
paramount: ‘Contract Administration’, ‘Dispute Resolution and Litigation’, ‘Planning
Code Compliance’, ‘Procurement Guidelines’, ‘Project Management’ and
‘Correspondence Analysis’. To effectively support these features, the nexus must be
underpinned by vector databases capable of storing essential legal and regulatory data.
Furthermore, a blueprint that seamlessly integrates suitable vector databases is paramount
to addressing the concerns raised by R4 and R14 in assessing the need for an NLP-PCMN.
The primary requirement for ‘Contract Administration’ is enabling users to access
pertinent contractual provisions governing specific scenarios, as R1, R18, and R19
emphasised. This process involves analysing contract documents, including clauses, legal
requirements and particular conditions. Similarly, incorporating a case law library
involves creating a classification system, metadata tagging and indexing. This integration
facilitates informed decision-making in the contract administration by consolidating all
required information. A case law library with databases for contract law provisions and
dispute resolution procedures vector database will formulate a ‘Dispute Resolution and
Litigation’ feature as proposed by R11.
Procurement guidelines are often subject to updates by government agencies, and they
present a challenge in tracking and incorporating the latest amendments. As highlighted
by R2, the ideal solution should furnish users with answers based on the most recent or
recent past guidelines, identifying and integrating the latest changes. A similar
mechanism can indeed be implemented for ‘Correspondence Analysis’ within the NLP-
PCMN framework. Various records such as letters, requests for information documents,
meeting minutes, progress reports, and instructions can be stored systematically, akin to
the approach outlined previously. This repository of documents helps project managers
stay up to date on all project-related correspondence.
Furthermore, statutes and ordinances pertinent to environment law protocols, labour law
provisions, and health and safety guidelines can be made accessible through vector
databases, enabling project management professionals to access this critical information
swiftly. Building code compliance is another crucial regulatory aspect that project
managers must adhere to. This can be achieved through specialised LLMs.
Blueprint for a natural language processing powered nexus for regulatory and legal landscape in
construction
Proceedings The 12th World Construction Symposium | August 2024 313
As articulated by R20, planning codes exhibit variations and are often dispersed across
different urban councils. The ‘Planning Code Compliance’ solution involves storing these
guidelines in a database and referencing attributes such as building type, location, height
requirements and diverse compliance constraints for building elements. Subsequently,
users can query the database by providing these attributes to ascertain the appropriate
constraints governing the design of a building on a vacant plot. This approach streamlines
the process of accessing and navigating planning codes during the design phase of
construction projects.
4.4 IMPLEMENTATION BLUEPRINT OF THE NLP - PCMN
To implement the NLP-PCMN, in the focus group stage, the R4, R5, R9 and R16
concluded the following blueprint for its architecture. The dash lines represent the vector
databases that provide sources for the specific LLM, such as ‘Contract Administration’.
All those primary services are connected, representing a nexus that provides an all-in-one
solution for legal and regulatory compliance management, as presented in Figure 2.
Figure 2: Blueprint for the NLP-PCMN (Source: Developed by authors)
A multi-step approach should be designed for ‘Contract Administration’, where a user
input and pre-existing database are used. The process starts by analysing the contract
P.V.I.N. Saparamadu, H.S. Jayasena, and B.A.I. Eranga
Proceedings The 12th World Construction Symposium | August 2024 314
document and identifying clauses, particular conditions, and contract data through
keyword detection and a domain-specific LLM. Then, the document is segmented and
converted into vector embeddings for efficient retrieval. When a user asks a question, the
system utilises an information retrieval algorithm to extract relevant vectors. The
retrieved vectors are then fed into an LLM specifically trained for contract interpretations,
generating a comprehensive response that includes the answer, related clauses and
document references. Figure 3 illustrates the above process.
Figure 3: Synthesising user input data with pre-existing databases (Source: Developed by authors)
For ‘Correspondence Analysis’ and ‘Procurement Guidelines Compliance’, it is vital that
up-to-date information is updated within the database. The following paragraphs describe
the implementation of such a system as proposed by R4. An information retrieval
algorithm segments the guidelines into individual clauses and extracts crucial information
such as the procurement guideline and the year. Incorporating new guidelines requires
adding them as new rows in the vector database. When users pose questions, the query
undergoes embedding using the same mechanism as Figure 3. Following the retrieval of
relevant clauses, a prompt is formulated utilising the retrieved information and the user’s
question. A check is conducted to identify any duplicate clauses for the same clause
number, if applicable, indicating them as “Previous” and “New” within the prompt.
Subsequently, a comprehensive answer is generated.
After analysing the literature alongside this study’s findings, several key points emerge.
The models discussed address specific tasks within contract management or the legal
landscape. In contrast, the proposed architecture aims to integrate all necessary functions
for managing legal and contractual aspects in the construction industry into one
comprehensive nexus. This holistic approach ensures that various tasks are handled
within a single, integrated system. Nevertheless, this study pioneers a consolidated NLP-
PCMN architecture. Moreover, this architecture leverages the capabilities of LLMs,
representing the forefront of NLP technology (Cambria & White, 2014). By using LLMs,
the system can provide more accurate and context-aware insights, making it a powerful
tool for construction’s legal and contract landscape.
Furthermore, the NLP-powered models that were introduced in the literature are tailored
to particular organisations. As concluded by R1, a publicly accessible NLP-PCMN can
democratise access to legal information. As Mitchell and Mancoridis (2006) highlighted
Blueprint for a natural language processing powered nexus for regulatory and legal landscape in
construction
Proceedings The 12th World Construction Symposium | August 2024 315
this architecture offers significant advantages. This enables individuals and businesses to
navigate the law without expensive legal consultations (Mitchell & Mancoridis, 2006).
5. CONCLUSIONS AND RECOMMENDATIONS
This study was directed to develop a blueprint for implementing an NLP-PCMN in the
construction industry. The experts that were interviewed all highlighted the potential use
cases and the need to implement an NLP-PCMN. One of the key considerations was the
complexity and vast amount of textual data in construction. Therefore, effective
construction management necessitates efficient storage and management of vast textual
data, a task beyond human capacity alone. Automation through NLP-PCMN not only
addresses this challenge by alleviating reliance on human memory but also empowers
stakeholders across various levels of expertise. Stakeholder empowerment through NLP-
PCMN extends to project managers, entry-level professionals, legal experts, and clients
alike. By serving as a legal handbook accessible during meetings and enhancing the
quality of life for industry newcomers, the NLP-PCMN promises to revolutionise
workflows and decision-making processes.
The implementation blueprint for NLP-PCMN, carefully crafted by expert focus groups,
underscores six significant features required to realise its full potential i.e. (i) ‘Contract
Administration’, (ii) ‘Dispute Resolution and Litigation’, (iii) ‘Planning Code
Compliance’, (iv) ‘Procurement Guidelines’, (v) ‘Project Management’ and (vi)
‘Correspondence Analysis’. The nexus is powered by the integration of domain-specific
vector databases. Those databases are formed by segmenting different documents in the
construction industry through keyword detection and metadata tagging and then
converted into vector embeddings for efficient retrieval. Furthermore, by facilitating the
constantly updating legal framework, the nexus model minimises the risk of non-
conformance.
Overall, by analysing all these findings, the NLP-PCMN represents a significant
advancement in the construction industry, offering a comprehensive solution that
addresses the intricate challenges faced by professionals in the legal and regulatory
landscape. Furthermore, it could lay the foundation for enhanced collaboration,
compliance and decision-making in the future of the construction industry.
To enhance the effectiveness and relevance of the NLP-PCMN, several key
recommendations are proposed. Firstly, the research findings emphasise the need for
NLP-powered tools in the construction industry. Therefore, it is recommended that
construction industry practitioners invest in integrating NLP-PCMN systems into their
existing management frameworks. Ideally, government agencies and professional
institutes should develop a publicly accessible NLP-PCMN system to ensure reliability
and widespread use. For research purposes, user-centric and publicly available NLP
models should be developed. Developers should use the requirements of the construction
industry, as identified in this study, to create effective solutions for the legal and
contractual management landscape in the construction industry.
This research contributes to academia by filling the gap for a consolidated architecture of
an NLP-PCMN. Additionally, it pioneers the concept of a consolidated NLP solution to
the construction industry as an NLP-PCMN. It is also important to acknowledge the
limitations of this study. This research considers the technological advancements up to
April 2024. Grounded in interpretivism, the study acknowledges the subjectivity inherent
P.V.I.N. Saparamadu, H.S. Jayasena, and B.A.I. Eranga
Proceedings The 12th World Construction Symposium | August 2024 316
in qualitative research, particularly when combining the knowledge from NLP developers
and construction practitioners. These limitations highlight the need for future studies to
consider broader samples and employ rigorous methods for interpreting qualitative data.
Furthermore, Future research should include a case study using a developed model to
identify potential time and cost savings as well as studies on technology adoption of NLP-
powered models in the construction industry.
6. REFERENCES
360iResearch (2024). Global AI in construction market by offering (services, solutions), deployment type
(cloud, on-premises), organisation size, application, industry type: Forecast 2024-2030 (Report
No. 5789100). United States. G. M. I. Inc.
https://www.researchandmarkets.com/report/construction#:~:text=The%20AI%20in%20Constru
ction%20Market,USD%203%2C226.92%20million%20by%202030
Babatunde, I., Ezirim, A., & Bid, O. (2023, September 3-9). Enhancing contract management through
Natural Language Processing(NLP): A case study of three African countries [Paper presentation].
5th Deep Learning Indaba, Accra, Ghana. https://openreview.net/forum?id=DUrRx78LEw
Beach, T. H., Hippolyte, J. L., & Rezgui, Y. (2020). Towards the adoption of automated regulatory
compliance checking in the built environment. Automation in Construction, 118(12).
https://doi.org/10.1016/j.autcon.2020.103285
Beach, T. H., Rezgui, Y., Li, H., & Kasim, T. (2015). A rule-based semantic approach for automated
regulatory compliance in the construction sector. Expert Systems with Applications, 42(12), 5219-
5231. https://doi.org/10.1016/j.eswa.2015.02.029
Cambria, E., & White, B. (2014). Jumping NLP curves: A review of natural language processing research.
IEEE Computational Intelligence Magazine, 9(2), 48-57.
https://doi.org/10.1109/MCI.2014.2307227
Hand, D. J., Christen, P., & Kirielle, N. (2021). F*: An interpretable transformation of the F-measure.
Machine Learning, 110(3), 451-456. https://doi.org/10.1007/s10994-021-05964-1
Hassan, F., & Le, T. (2020). Automated requirements identification from construction contract documents
using natural language processing. Journal of Legal Affairs and Dispute Resolution in Engineering
and Construction, 12(2). https://doi.org/10.1061/(ASCE)LA.1943-4170.0000379
Jallan, Y., Brogan, E., Ashuri, B., & Clevenger, C. M. (2019). Application of natural language processing
and text mining to identify patterns in construction-defect litigation cases. Journal of Legal Affairs
and Dispute Resolution in Engineering and Construction, 11(4).
https://doi.org/10.1061/(ASCE)LA.1943-4170.0000308
Kelley, G. (2012). Construction law: An introduction for engineers, architects, and contractors. Wiley &
Sons. http://ndl.ethernet.edu.et/bitstream/123456789/13047/1/116.pdf
Khurana, D., Koli, A., Khatter, K., & Singh, S. (2023). Natural language processing: State of the art, current
trends and challenges. Multimedia Tools and Applications, 82(3), 3713-3744.
https://doi.org/10.1007/s11042-022-13428-4
Lauriola, I., Lavelli, A., & Aiolli, F. (2022). An introduction to deep learning in natural language
processing: Models, techniques, and tools. Neurocomputing, 470, 443-456.
https://doi.org/10.1016/j.neucom.2021.05.103
Lee, J., Ham, Y., Yi, J.-S., & Son, J. (2020). Effective risk positioning through automated identification of
missing contract conditions from the contractor’s perspective based on FIDIC contract cases.
Journal of Management in Engineering, 36(3). https://doi.org/10.1061/(ASCE)ME.1943-
5479.0000757
Lee, J.-K. et al. (2023). High-level implementable methods for automated building code compliance
checking. Developments in the Built Environment, 15(12).
https://doi.org/10.1016/j.dibe.2023.100174
Lee, J., Yi, J.-S., & Son, J. (2019). Development of automatic-extraction model of poisonous clauses in
international construction contracts using rule-based NLP. Journal of Computing in Civil
Engineering, 33(3). https://doi.org/10.1061/(ASCE)CP.1943-5487.0000807
Blueprint for a natural language processing powered nexus for regulatory and legal landscape in
construction
Proceedings The 12th World Construction Symposium | August 2024 317
Madan, R., & Ashok, M. (2023). AI adoption and diffusion in public administration: A systematic literature
review and future research agenda. Government Information Quarterly, 40(1).
https://doi.org/10.1016/j.giq.2023.101881
Marzouk, M., & Enaba, M. (2019). Text analytics to analyse and monitor construction project contract and
correspondence. Automation in Construction, 98, 265-274.
https://doi.org/10.1016/j.autcon.2018.11.018
Mitchell, B. S., & Mancoridis, S. (2006). On the automatic modularisation of software systems using the
bunch tool. IEEE Transactions on Software Engineering, 32(3), 193-208.
https://doi.org/10.1109/TSE.2006.31
Moon, S., Lee, G., & Chi, S. (2022). Automated system for construction specification review using natural
language processing. Advanced Engineering Informatics, 51.
https://doi.org/10.1016/j.aei.2021.101495
Murdoch, J., & Hughes, W. (2002). Construction contracts: Law and management (5th ed.). Routledge.
https://www.nibmehub.com/opac-
service/pdf/read/Construction%20Contracts%20Law%20and%20Management%205th%20Editio
n.pdf
Nadkarni, P. M., Ohno-Machado, L., & Chapman, W. W. (2011). Natural language processing: An
introduction. Journal of the American Medical Informatics Association, 18(5), 544-551.
https://doi.org/10.1136/amiajnl-2011-000464
Padhy, J., Jagannathan, M., & Delhi, V. (2021). Application of natural language processing to automatically
identify exculpatory clauses in construction contracts. Journal of Legal Affairs and Dispute
Resolution in Engineering and Construction, 13(4). https://doi.org/10.1061/(ASCE)LA.1943-
4170.0000505
Parikh, P. M., Shah, D. M., & Parikh, K. P. (2023). Judge Juan Manuel Padilla Garcia, ChatGPT, and a
controversial medicolegal milestone. Indian Journal of Medical Sciences, 75(1), 3-8.
https://doi.org/10.25259/IJMS_31_2023
Potrac, P., Jones, R. L., & Nelson, L. (2014). Interpretivism. In L. Nelson, R. Groom & P. Potrac (Eds.),
Research Methods in Sports Coaching (pp. 31-41). Routledge.
https://doi.org/10.4324/9780203797549-4
Saunders, B., Sim, J., Kingstone, T., Baker, S., Waterfield, J., Bartlam, B., Burroughs, H., & Jinks, C.
(2018). Saturation in qualitative research: Exploring its conceptualisation and operationalisation.
Quality & Quantity, 52(4), 1893-1907. https://doi.org/10.1007/s11135-017-0574-8
Shaikh, A., & Gohar, S. (2024, May 30-31). Study on use of text based AI Tool in construction industry
Contracts [Paper Presentation]. 3rd IBA SBS International Conference 2024, Karachi, Pakistan.
https://ir.iba.edu.pk/sbsic/2024/program/59
Soibelman, L., Wu, J., Caldas, C., Brilakis, I., & Lin, K.Y. (2008). Management and analysis of
unstructured construction data types. Advanced Engineering Informatics, 22(1), 15-27.
https://doi.org/10.1016/j.aei.2007.08.011
Stone, P., Brooks, R., Brynjolfsson, E., Calo, R., Etzioni, O., Hager, G., Hirschberg, J., Kalyanakrishnan,
S., Kamar, E., & Kraus, S. (2021). Artificial intelligence and life in 2030: The one hundred year
study on artificial intelligence. United States. University of Stanford.
https://doi.org/10.48550/arXiv.2211.06318
Szewc, T. (2022). The impact of legal regulations on investment project management in construction.
Organisation and Management: Scientific Quarterly, 1(5), 145-154.
https://doi.org/10.29119/1899-6116.2021.57.10
Wu, C., Li, X., Guo, Y., Wang, J., Ren, Z., Wang, M., & Yang, Z. (2022). Natural language processing for
smart construction: Current status and future directions. Automation in Construction, 134.
https://doi.org/10.1016/j.autcon.2021.104059
Yan, H., Yang, N., Peng, Y., & Ren, Y. (2020). Data mining in the construction industry: Present status,
opportunities, and future trends. Automation in Construction, 119(1).
https://doi.org/10.1016/j.autcon.2020.103331
Zhang, D., Mishra, S., Brynjolfsson, E., Etchemendy, J., Ganguli, D., Grosz, B., Lyons, T., Manyika, J.,
Juan, Sellitto, M., Shoham, Y., Clark, J., & Perrault, R. (2022). The AI index 2022 annual report.
Stanford University. https://doi.org/10.48550/arXiv.2205.03468
