Knowledge-Based Systems 86 (2015) 194–209

Contents lists available at ScienceDirect

Knowledge-Based Systems
journal homepage: www.elsevier.com/locate/knosys

A knowledge-based expert system for assessing the performance level

of green buildings
Mehrbakhsh Nilashi b,⇑, Rozana Zakaria a, Othman Ibrahim b, Muhd Zaimi Abd. Majid e, Rosli Mohamad Zin a,
Muhammad Waseem Chugtai c, Nur Izieadiana Zainal Abidin c, Shaza Rina Sahamir c, Dodo Aminu Yakubu d
a
Construction Research Alliance, Universiti Teknologi Malaysia, Johor Bahru, Johor, Malaysia
b
Faculty of Computing, Universiti Teknologi Malaysia, Johor, Malaysia
c
Faculty of Civil Engineering, Universiti Teknologi Malaysia, Johor, Malaysia
d
Faculty of Built Environment, Universiti Teknologi Malaysia, Johor, Malaysia
e
Institute for smart infrastructure and innovative construction, 81310, UTM Johor Bahru, Johor, Malaysia

a r t i c l e i n f o a b s t r a c t

Article history: Sustainability has become an important initiative discussed and undertaken, not only by private
Received 17 January 2015 buildings, but also by public buildings which both dealing with residential, office, commercial as well
Received in revised form 2 June 2015 as hospital. Sustainable building is the practice of designing, constructing, operating, maintaining, and
Accepted 8 June 2015
removing buildings in ways that conserve natural resources and reduce pollution. Rating systems provide
Available online 14 June 2015
effective framework for assessing building environmental performance and they measure a building’s
sustainability by applying a set of criteria organized in different categories. A good Green Building
Keywords:
Rating System (GBRS) should cover key indicators reflecting a building’s characteristics and keep their
AHP
Fuzzy inferences system
performance in balance. This paper proposed a knowledge-based expert system as a tool to assess the
Fuzzy rule performance level of a green building based on assessment factors of green building rating systems.
Green building Analytic Hierarchy Process (AHP) and fuzzy logic is adopted in order to develop the knowledge-based
Rating system expert system. The data for this research collected from the experts in the field via pair-wise and
Likert-based questionnaires. Using AHP, the most important parameters of rating systems according to
their weights selected to be incorporated in the Fuzzy Inferences System (FIS) of fuzzy logic model.
The fuzzy rules (knowledge) discovered from the collected data for FIS to assess the performance level
of the green buildings from the Environmental, Social and Economical perspectives denoted as SE2.
The outcome of this research is accordingly a performance assessment tool that analyzes the effect of
factors in developing the sustainable building.
Ó 2015 Elsevier B.V. All rights reserved.

1. Introduction environmental issues. The concept emphasizes the integration of

Sustainability initiatives are considerable in the erection of new
buildings, in the operation and the renovation of existing buildings
and nowadays it is apparent that it has important impact on the
building industry. As one of key outputs of the construction indus-
try, buildings largely reflect resources and waste impacts during its
lifecycle. Construction activities in general known as resource
intensive and the impacts reflected in consumption of natural
resources and pollutions [34].
Generally, sustainability is the ability of a system to continue on
an indefinite basis typically incorporating economic, social, and
⇑ Corresponding author.
E-mail address: nilashidotnet@hotmail.com (M. Nilashi).
humans in nature and requires that human activity remain within
bounds avoiding impact on ecological systems [36]. Achieving a
low carbon economy through building usage is necessary as it is
responsible for approximately 40% of CO2 emissions in the UK
and across the EU, therefore policies on tackling energy use
through design and development is of priority [37].
The awareness and importance of maintaining sustainable
developments within the planning and engineering sector has led
many to look toward new and innovative ways to incorporate sus-
tainability into their designs. The term ‘‘green’’ building defines
environmentally friendly techniques and technologies used in the
design and construction of the built environment [38]. The green
building revolution is sweeping across not only the United States
but also most of the world. This revolution is further fueled by

http://dx.doi.org/10.1016/j.knosys.2015.06.009
0950-7051/Ó 2015 Elsevier B.V. All rights reserved.
 M. Nilashi et al. / Knowledge-Based Systems 86 (2015) 194–209
the knowledge that the world has little time to respond to the
growing dangers of climate change, especially global warming,
and that buildings play a huge role in causing carbon dioxide emis-
sions that drive global climate change [73]. Green building tech-
nology for responsiveness to sustainable development is not only
for private buildings but also for public buildings which dealing
with residential, office, commercial as well as hospital has become
a flagship of sustainable development in this century. Its goal takes
the responsibility for balancing long-term economic, environmen-
tal and social health [71,58].
Assessing the performance of green buildings is a critical task as
the different rating systems emphasize different aspects of build-
ing performance and has been an eye-catching topic these days
[61,29,13]. Due to the fact that Men strive to increased comfort
and financial independence, the effects of economic, and the qual-
ity of life is being hampered. While at the same time are also neg-
ative health effects that must bear from the damage due to the
rising number of environmental catastrophes as result causes cli-
matic change is experienced [48]. Sustainable building design will
become a more common practice once interior environment pri-
marily the productivity gains believed to be associated with the
provision of high quality interior environments [24]. For example,
research on environmentally friendly design and management sys-
tems has being conduct over the life cycle of buildings [53,63].
Besides, there are various methods for assessing the performance
of buildings. Many of these methods emphasize the impact of
buildings on the global environment and individual health, focus-
ing on energy use, indoor climate and other environmental issues
[18].
MCDM or MCDA are well-known acronyms for multiple-criteria
decision-making and multiple-criteria decision analysis. MCDM is
concerned with structuring and solving decision and planning
problems involving multiple criteria [40]. It has been utilized to
solve many real-world problems [35,64,60,28]. Zadeh [74] intro-
duced the concept of fuzzy sets to enable analysts to deal with
imprecise and subjective concepts and to deal with linguistic vari-
ables in various decision and evaluation applications. Fuzzy MCDM
(FMCDM) problems [26,65,41,39,12,10], among which the ratings
and the weights of criteria evaluated on imprecision, uncertainty,
and vagueness are usually expressed by linguistic terms and then
set into fuzzy numbers.
A research by Lu et al. [39] that integrated human actions and
complex socio-economic themes into the process of New Product
Development (NPD) in order to adapt its design to various
competitive markets showed the effectiveness of Fuzzy
Multi-Criteria Decision Making (FMCDM) in theme-based product
evaluation. Their method combined MCDM with Group Decision
Making (GDM) methods and proposed hierarchical operators to
fuse the data obtained from both human evaluators and
machines.
We also take the advantage of MCDM and fuzzy set theory and
develop an assessment tool for green buildings performance. In our
proposed method, Analytic Hierarchy Process (AHP) is used for
group decision making and ranking the performance assessment
criteria and fuzzy inference system is applied for final performance
evaluation of green buildings. Hence, in comparison with research
efforts found in the literature, our work has the following differ-
ences. In this research:
1. Using AHP, the performance assessment criteria for green build-
ings are ranked and weighted from the SE2 perspectives.
2. Using fuzzy set theory, a new knowledge-based expert system
for assessing the performance level of green buildings from
the Environmental, Social and Economical perspectives is pro-
posed.
195
1.1. Benefits of green buildings
Implementation of green building practices is believed will pos-
sibly achieve the three benefits, which is SE2 benefits [70]. The
environmental benefits include: Biodiversity and ecosystems
enhancement; Air and water quality improvement; Waste streams
reduction; Natural resources conservation and restoration; and
Minimizing the global warming.
While secondly, the economic benefits contribute to reduction
in operating and maintenance costs; Green product and services
creation, expansion and markets shape; Occupant productivity
improvement; Occupant absenteeism minimization; Life-cycle
economic performance optimization; Building image improve-
ment; and reduce the civil infrastructure costs.
On the other hand, social benefits give to reduction in operating
and maintenance costs; Green product and services creation,
expansion and markets shape; Occupant productivity improve-
ment; Occupant absenteeism minimization; Life-cycle economic
performance optimization; Building image improvement; and
reduce the civil infrastructure costs.
In a green building, energy efficiency used to describe in fulfill
several criteria which need to be achieve. This include the use of
energy efficient equipment, suitability of materials for the climate
conditions, the service and amenities provided must fulfill the
building use and the building should consumed less energy than
the similar buildings. Besides, another important aspect should
be considered is the embodied energy in both building construc-
tion and demolition [47]. Many countries have introduced to
reduce the building energy in order to improve the energy effi-
ciency in the building sector.
Since sustainable development with Social, Environmental, and
Economical (SE2) principles encouraged to response to the mea-
surement of carbon footprints, many rating systems were and
being developed in order to assess the ‘‘green-ness’’ of green build-
ings and many in the field are thinking about measuring the actual
performance of them i.e. ‘‘ the performance of buildings that we
considered green.’’ For instance, many countries have developed
their own rating systems either by setting up their own parameters
or by modifying from the rating tools developed in other countries.
Green Building Rating Systems (GBRS) is one of such systems that
measure a building’s sustainability by applying a set of criteria
organized in different categories [11]. In these GBRS, the criteria
perform various functions in measuring responsiveness toward
the sustainable development. GBRS can support the
decision-making process and increase the efficiency of actions by
simplifying, clarifying and making aggregated information
available.
1.2. The problem and our contribution
Assessing the performance of green buildings in their
post-occupancy stage is the most effective way to insure that a
building obtained what degree of success in its design. As the green
building approach should consider three main criteria are Social,
Environmental, and Economical (SE2); hence, the assessment tool
for measuring the performance necessarily to take these criteria
into consideration [11].
Throughout the phases of building life span, environmentally
friendly built environments should be associated with safety, secu-
rity, wellbeing, convenience, reasonable cost and long-term adapt-
ability. Satisfaction of these criteria achieves an optimal
combination of SE2 values for buildings [14,15,32,66]. Social, eco-
nomic, environmental and technological dimensions are all impor-
tant in evaluating building performance [16]. A wide range of
criteria have been developed in order to assess building
196 M. Nilashi et al. / Knowledge-Based Systems 86 (2015) 194–209

performance; however, there is a lack of consensus on what factors input to an output using fuzzy logic, in a different and entirely
constitute excellence in this area. Recently, several quantitative new way for green building rating systems.
and qualitative approaches have emerged for assessing sustainabil- The remainder of this paper is organized as follows: In Section 2,
ity of buildings. the related work on green building rating systems is present. In
GBRS helps in evaluating the criteria performance in buildings Section 3, applied methodology for proposed expert system is
that influences the society, e.g., energy consumption, waste pro- explained fuzzy-AHP is introduced. Section 4 provides an empirical
duction and indoor air quality, in order to improve efficiency. It study. Finally, conclusions and future work is present in Section 5.
is use as a tool to track performance and provide building owners
and developers with a guide to assist in building more sustainably.
2. Related work on green building rating systems
Based on information from the Intergovernmental Panel on
Climate Change (IPCC), if the approach to constructing and operat-
Green building rating systems measure a building’s sustainabil-
ing buildings remains as the status quo, there will be major eco-
ity by applying a set of criteria organized in different categories,
nomic and environmental repercussions [6,11]. GBRS provide a
such as ‘‘Site selection’’, ‘‘Energy’’, ‘‘Water’’, ‘‘Resources’’,
way for continually updating practices and procedures to ensure
‘‘Material and Components’’, ‘‘Environmental’’, ‘‘Loadings’’,
continual improvement and innovation [6,11].
‘‘Transport’’, ‘‘Emissions’’, and ‘‘Waste’’. For each criterion, a certain
GBRS measure a building’s sustainability by applying a set of
number of points is allocated. The total score defines the type of
criteria, which organized in different categories and classified fac-
certification a building receives.
tors to conduct the factor wise performance evaluation of build-
Sustainability labeling or credit incentive programs are often
ings. The existing models developed in the previous studies [46]
easily understandable and make it possible to obtain certain levels
for the rating systems adopt mostly quantitative methods, so there
of ‘‘sustainability’’ or ‘‘greenness’’. Sustainability credit incentive
is a lack of qualitative evaluation regarding user’s interaction with
programs, in examples, provide tools and performance criteria for
green buildings. In general, research on green building rating sys-
the implementation and recognition of sustainable practices.
tems can be divided into two main areas: (1) identification of cri-
Credit programs are effective on two levels: they provide specific
teria for the development of rating systems, and (2) research area
criteria for achieving various credits, and thus give clear direction
concerns on assessment and validation of green building rating
for the implementation of sustainable practices; and they provide
systems. Such studies mostly emphasize energy performance in
recognition for credit achievement, and thus, provide an incentive
buildings and generally adopt quantitative methods. This is the
for the implementation of sustainable practices.
reason why there is a lack of qualitative evaluation methods in
When designing a green/sustainable building, it is important to
term of exploring user experience of buildings.
ensure not only that criteria are being met, but also that the build-
In order for a GBRS to be widely recognized, it is critical that the
ing will perform as intended. That is, the methods implemented for
system reflects and understanding to the needs of end-users
achieving credits should result in true progress toward achieving a
[13,29,30]. By contrast, in this study, a fuzzy inference system
balance between economy, ecology, efficiency, and functionality, in
employing fuzzy decision making rules applied to model the qual-
a socially responsible manner.
itative aspects of human knowledge and reasoning process without
Many related research studies about establishing the proper set
employing precise quantitative analyses.
of green building criteria have been undertaken in the last decade,
The questions that have been raised for this study are: (a)
with an outlook for additional necessary investigation. The litera-
Which factors are important to assessing the performance of green
ture review gives a summarized overview of the state-of-art and
building on the three main criteria of SE2 perspectives? (b) How
knowledge about which criteria a green building rating system
AHP and fuzzy logic methods being employed for assessing the
has to consider in order to properly measuring a building’s sustain-
green buildings performance?
ability. Different political and voluntary organizations, alliances of
Hence, this paper discussed an efficient expert system devel-
industry and academia, and individual and groups of scientists
oped for performance assessment of a green building system with
have conducted researches in this field.
adopting Analytic Hierarchy Process (AHP) and fuzzy logic denoted
Generally, two types of assessment tools have been developed
as (fuzzy-AHP). Using these approaches, the system is constructed
by building sector which the first group of them is tools which
based on three main assessment dimensions in green buildings are
includes the criteria into the system for measuring the greenness
Social, Environmental, and Economical denoted as SE2. These three
level. However, in the second group Life Cycle Assessment (LCA)
main dimensions have been widely used in the most of the current
methodology is considered. In addition, the assessment systems
rating systems. The main objectives of this research are twofold:
based on LCA have aimed to be used for selection of building mate-
rial, building design, and local utility options such as transport
1. Evaluating and weighting the criteria for green buildings from
type, energy supply and waste management during the design
the SE2 perspectives.
phase [6]. Examples of tools of this category that contain LCA com-
2. Developing a new knowledge-based expert system for assessing
ponent are KCL Eco (Finland)-KCL (2005), Beat (Denmark)-DBRI
the performance level of green buildings from the SE2
(2005), Bees (USA)-OAE (2004) and EcoQuantum (Netherlands).
perspectives.
In contrast with the LCA-based assessment schemes, in the
criteria-based approaches, for each criteria specific scale ranging
In addition, the main contribution of the paper is an integrated
for example ‘‘small’’ and ‘‘large’’ environmental impact is consid-
fuzzy-AHP Logic approach for assessing the performance level of
ered which these point values assess a selected number of param-
green buildings from the SE2 perspectives. Moreover, using fuzzy
eters. The criteria-based rating systems have been used as
logic, the fuzziness and nonlinearity of the impact of SE2 factors
comprehensive environmental assessment tools in the worldwide.
on performance are address properly. To the best knowledge of
In the following some of GBRS are introduced [21,6,11].
the authors, there are few empirical studies to assess the perfor-
mance of green building using fuzzy sets theory and the majority
of the previous researches are based on fuzzy multi-criteria based 2.1. GBTool (Canada)
methods such as fuzzy AHP. However, for the first time this
research work employed AHP and Fuzzy Inferences Systems (FIS), In 1996, the Green Building Challenge (GBC) developed the
which is the process of formulating the mapping from a given GBtool in Canada (Greenman Sustainable Buildings – Green
 M. Nilashi et al. / Knowledge-Based Systems 86 (2015) 194–209
Building, Consulting, Education and Sales in Canada and USA). It is
a customizable building rating system, which evaluates environ-
mental and sustainability performance. The system is designed as
a generic toolbox, which can be customized according to local
and regional building performance requirements and needs.
GBTool uses a scoring system based on scale of 1 (deficient), 0
(minimum pass), +3 (good practice), and +5 (best practice).
2.2. CASBEE
CASBEE is Japan’s version of a green building certification pro-
cess and has been in development since 2001. The website asserts
that it is promoted under the Kyoto Protocol, an international pro-
tocol to the United Nations Convention on Climate Change, as a
tool to reduce emissions. As per the CASBEE website, the two
assessment categories for CASBEE are Building Environmental
Quality, and Performance and Reduction of Building
Environmental Loadings. The CASBEE system arrives at a score
from a formula that is derived from the form that the client fills
out for their building. Scores range from poor (C) to excellent (S)
and buildings are assigned one of the 5 different classification let-
ters, C, B, B+, A and S. Uniquely, this system rates all levels of
building performance from best to worst, instead of praising the
good alone. CASBEE has four grading categories (i) pre-design, (ii)
new construction, (iii) existing buildings, and (iv) renovations,
which are evaluated based on five criteria: (a) Energy Efficiency;
(b) Site Selection; (c) Indoor environmental quality; (d)
Resources and materials, and (e) Water conservation.
2.3. BREEAM
A very popular British program of the Building Research
Establishment in the UK is BREEAM and the world was eventually
introduced to BREEAM International. Building project managers
work with a BREEAM Assessment Organization from the beginning
of the planning phases. As indicated by their website [7], the
method assesses buildings against a set criteria and provides an
overall score which will fall within a range providing either a
Pass, Good, Very Good, Excellent or Outstanding rating.
USGBC Similar to BREEAM, the USGBC has a green building rat-
ing process called LEED, which is one of the industry standards in
the US and abroad. As stated by the USGBC website, the LEED rat-
ing and certification program has been in existence since the year
1999. Projects are required to have a LEED Accredited Professional
(AP), much like the BREEAM program, in order to register a project
and obtain any of the four rating levels. The LEED rating system
includes four levels of certification: Certification for meeting the
prerequisites and reaching 26–32 points; Silver for achieving the
next level of points: 33–38; Gold for achieving an even higher level
of points: 39–51; and Platinum for meeting or exceeding all criteria
for certification: 52–69 points. A site or project earns one of these
different certification levels by achieving points listed for a variety
of credits under six categories: Sustainable Sites, Water Efficiency,
Energy and Atmosphere, Materials and Resources, Indoor
Environmental Quality (IEQ), and Innovation and Design Process.
2.4. DGNB
The German Sustainable Building Certificate proposes that
building sustainably means to build intelligently [23]. ‘‘The focus
is on comprehensive quality concept that serves the building and
real estate sectors, as well as society in general. Sustainable prop-
erties are beneficial to the environment, conserve resources, com-
fortable and healthy for their users, and fit optimally into their
socio-cultural surroundings’’ [17]. The German sustainable rating
system was developed through a partnership by the German
197
Sustainable Building Council (DGNB) and the Federal Ministry of
Transport, Building, and Urban Affairs (BMVBS), and it was
released in 2008. It was created to response to the changing cli-
mate of the German real estate market, and to keep German real
estate competitive and attractive to potential investors [17]. It is
a voluntary based system covering six categories of evaluation:
Ecology, Economy, Social-cultural and functional topics,
Techniques, Processes, and Location. Based on the number of
points awarded in each category, the buildings are then awarded
with a rating of bronze, silver, or gold. Buildings are awarded a
pre-certificate rating based on their projected construction and
design of the building.
2.5. BEAM
In 2002, the Business Environment Council (BEC), and HK-Beam
Society developed BEAM in Hong Kong HK BEAM Society. It evalu-
ates and measures the environmental performance of buildings in
Hong Kong. The evaluation is based on five building performance
criteria: Hygiene, health, comfort, and amenity, Land use, site
impact, and transportation, Use of materials, recycling, and waste
management, Water quality, conservation, and recycling; and,
Energy efficiency, conservation, and management.
2.6. Green Globes
The Green Building Institute’s (GBI’s) Green Globes can
offer specific advantages in terms of affordability and user-
friendliness, making it well-suited for smaller, financially limited
projects that want to quantify green construction and operation
[22].
2.7. Green Star
Green Star is a voluntary environmental rating system for build-
ings in Australia. It was launched in 2003 by the Green Building
Council of Australia. Nine categories are assessed with the Green
Star tools: Management, Indoor environment quality, Energy,
Transport, Water, Materials, Land use & ecology, Emissions and
Innovation [50].
2.8. LEED
According to Yanarella et al. [69], LEED and other green rating
systems are considered ‘‘transitional sustainability’’, although still
moving in the direction to achieving real sustainability. Rating sys-
tems are given this position on the scale for two reasons, the first is
the risk associated with attempting to place the idea of sustainabil-
ity onto a single building, without taking a holistic approach. The
second reason is the fact that changes in governance can derail
any advances in this area of sustainability [69].
From the above description of rating systems, it can be seen that
each of the rating systems allocates marks or scores to various
aspects of sustainability. If these aspects are grouped into domains,
and the maximum possible marks for the aspects aggregated, we
can arrive at the aggregate maximum marks for each domain and
hence the relative weight assigned to each domain by the rating
system. A comparison of such relative weights used in the Pacific
Northwest was made by Fowler and Rauch [19] and is summarized
in Table 1.
According to Sahamir and Zakaria [58], Green Building Index
(GBI) [20] which is presented in Table 2 is Malaysia’s first compre-
hensive green rating system for buildings and towns, created to
promote sustainability in the built-environment and raise aware-
ness of environmental issues among developers, architects, engi-
neers, planners, designers, contractors as well as the public. GBI
198 M. Nilashi et al. / Knowledge-Based Systems 86 (2015) 194–209

Table 1
Comparison of rating systems using WBDG principles [19].

System Weightage (%)

Site (%) Energy (%) Water (%) Materials (%) Enhance indoor Optimize operational and Other (%)
environmental quality (%) maintenance practice
BREEAM 15 25 05 10 15 15% 15
CASBEE 15 20 02 13 20 15% 15
Green Globes 11.5 36 10 10 20 – 12.5
LEED 20 25 07 19 22 – 07

Table 2 categories which summarized the building performance at the

Points distribution for each green rating criteria [58]. level of some key-sustainability aspects and 25 sustainability indi-
Green criteria/elements Weightage (%) cators within the three sustainability dimensions obtained from
GBI GBI BREEAM LEED G
study by Mateus and Braganca [45]. The assessment categories
NRNC NREB 100% 100% STAR included Climate change and outdoor air quality, Land use and bio-
100% 100% 100% diversity, Energy Efficiency, Materials and waste management,
Energy efficiency 35 38 17.27 35.45 16.86 Water efficiency, Occupant’s health and comfort, Accessibilities,
Indoor environmental 21 21 13.64 16.36 18.60 Education and awareness of sustainability, and Life-cycle costs.
Sustainable site 16 10 10.91 16.36 9.88 Each assessment category is identified by one or more indicators.
planning and
A survey carried out by [11] among building construction sector
management
Materials and resources 11 9 11.36 14.55 20.35 professionals in a particular national context sought relative
Water efficiency 10 12 5.45 8.18 8.14 weights for various domains and aspects for a rating system in
Innovation 7 10 9.09 5.45 2.91 Sri Lankan. They used direct ranking and Analytic Hierarchy
Transport 7.27 6.98 Process (AHP) methods for assessing the components of rating sys-
Land use and ecology 9.09 4.65
Pollution 9.09 11.63
tem. They considered six domains of Site, Energy efficiency, Water
Waste 6.82 efficiency, Materials, Indoor environmental quality and Waste and
Regional priority credits 3.64 Pollution for the rating system. The overall domain weights were
compared with those assigned in eight other rating systems, orig-
inating from eight different countries.
is developed specifically for the Malaysian-tropical climate, envi- By comparing the commonly used evaluation methods of green
ronmental and developmental context, cultural and social needs. buildings, [72] developed an assessment method for green store
The GBI is based upon the existing rating tools such as the buildings in China. The method references the rating requirements
Singapore Green Mark and the Australian Green Star system, set by the ‘‘China Green Building Evaluation Standard’’ and
among others which have been extensively modified for the weighted credits for each category. The Expert Group Decision
Malaysian application [56]. There are 10 versions of GBI rating sys- AHP method are used to develop the weighting system for green
tems; 1. Residential new construction (RNC), 2. Non-residential store buildings. The indicator system of the green store building
new construction (NRNC), 3. Non-residential existing building rating standard included the seven categories: Landscape, Energy
(NREB), 4. NRNC Data Centre, 5. NREB Data Centre, 6. Industrial efficiency, Water efficiency, Material and resources, Indoor envi-
new construction (INC), 7. industrial existing building (IEB), 8. ronment, Construction management, and Operation management.
NRNC retail, 9. NREB retail and 10. Township. By comprehensive The weight distributions highlighted the importance of Indoor
review of literature, a comparison among the rating systems has environmental quality, Energy efficiency and Operation manage-
been made by [58], which is presented in Table 2. The aim for their ment within store buildings.
study was to investigate green assessment criteria for public hos- Using a Delphi technique, Alyami et al. [5] found that the lead-
pital building development in Malaysia. It compiled the essential ing international sustainable assessment schemes, such as
criteria of existing green rating systems for healthcare buildings BREEAM and LEED, are inapplicable for the Saudi context.
worldwide and presents the difference between each criteria com- Accordingly, they developed a new rating system with further cat-
pared to Malaysian green rating system. Results from the analysis egories and criteria for the assessment of the built environment in
showed the important assessment criteria of green public hospital Saudi Arabia. The developed building environmental and sustain-
building correspond to Malaysia. able assessment categories included: Indoor environmental qual-
In many researches, in developing the rating systems, three ity, Energy efficiency, Water efficiency, Waste management, Site
dimensions of assessments SE2 for green building have been quality, Material, Pollution, Quality of services, Economic aspects
considered. Cultural aspects and Management and Innovation. Each of the
Ali and Al Nsairat [4] studied international green building above categories included a list of related criteria, creating a 92
assessment tools and then defined new assessment items respect- item list of criteria for sustainable residential building assessment
ing the local conditions of Jordan. They analyzed the main charac- in Saudi Arabia.
teristics of several building environmental performance Kabak et al. [28] developed a ‘‘fuzzy multi-criteria decision
assessment systems in different countries and studied the local making (MCDM)’’ approach in order to analyze National Building
context. In their developed rating system the researchers defined Energy Performance Calculation Methodology (BEP-TR). Their
seven main categories for the assessment tool which were Site, approach was applied to categorize alternative buildings according
Energy efficiency, Water efficiency, Material, Indoor environment to their overall energy performance. They discussed the results of
quality, Waste and pollution, and Cost and economics. their study in terms of developing a new and practical building rat-
Berardi [9] designed a rating system using these dimensions ing system.
and several indicators and categories (global indicators) and indi- Marzouk et al. [44] developed a rating system for green bridges
cators. The designed system includes a total of nine sustainability and Simos’ procedure was followed to acquire weights of criteria.
 M. Nilashi et al. / Knowledge-Based Systems 86 (2015) 194–209
In their research, a key-list of gathered criteria was retrieved from
the literature and discussed with bridge experts via un-structured
interviews. To elite the important criteria that affect the sustain-
ability of bridge projects, they developed a questionnaire survey
and finally Twenty-one criteria have been selected and considered
by the experts using the questionnaire survey.
Based on the analysis of the conventional building energy cer-
tificates, Koo et al. [31] developed a new energy efficiency rating
system for existing residential buildings from two perspectives:
(i) establishment of reasonable and fair criteria for the building
energy efficiency rating system; and (ii) establishment of compar-
ative incentive and penalty program to encourage the voluntary
participation of all residents in the energy saving campaign. In
their research, the multi-family housing complex was selected as
the representative type of existing residential building in South
Korea.
3. Applied methodology for proposed expert system
The results of the current literature survey showed that there is
no study that uses AHP and fuzzy Logic for performance assess-
ment of green buildings and the effect of SE2 factors on the overall
performance. This study is then conducted with a combined AHP
and Fuzzy logic approach as fuzzy-AHP for assessing the perfor-
mance of green buildings.
Our research aims to develop a method for evaluating green
building performance, which is mainly based on human knowledge
and experience. The knowledge-based expert system use human
knowledge to solve problems that normally would require human
intelligence [77,76]. In addition, the purpose of this paper is to pre-
sent an application of fuzzy logic for performance assessment of
green buildings. Accordingly, the proposed fuzzy-logic based per-
formance evaluation model uncovers some of the hidden
Fig. 1. Research methodology framework.
199
relationships between components of green building assessment
methods of SE2 along with their indicators and the level of perfor-
mance of green buildings. It should be noted that human experi-
ence about a green building plays an important role in
constructing the proposed expert system. Also, accuracy of the dis-
covered fuzzy rules in the fuzzy logic for FIS system is more
depends on human knowledge and experience. We conduct a case
study, in which we use Malaysian green buildings and the knowl-
edge (fuzzy rules) in the fuzzy logic system is discovered from the
collected data. A general overview of the research methodology for
this study is presented in Fig. 1.
3.1. Analytic Hierarchy Process (AHP)
AHP is a systematic procedure for dealing with decision making
problems with many alternatives. AHP is based on a hierarchical
structuring of decision making elements using pairwise compar-
isons. This technique is fairly simple, practical, and can be per-
formed using the steps as showed in Fig. 2 [57,27,52,3,2]. In AHP,
to assign judgment in comparing the pairs of alternatives in each
level of the hierarchy, a scale of 1–9 (1 – Equally Preferred; 5 –
Strongly Preferred; 9 – Extremely Preferred) is recommended
(see Table 3).
3.2. Fuzzy logic
Fuzzy logic evolved from the desire to train computer systems
with human expertise. Fuzzy set theory has been developed for
modeling complex systems under uncertain or imprecise environ-
ments [55,8]. It has been widely applied in decision-making and
problem solving [68,33,28,75,54,59,1]. Converting the knowledge
of the human expert to an equation that a computer can process
is difficult when the process involves a number of variables and
200 M. Nilashi et al. / Knowledge-Based Systems 86 (2015) 194–209

Table 4
Assessment dimensions, indicators and parameters for proposed model.

Dimension Indicators Parameters

Environmental Site selection Site design
Land use
Landform
Microclimate
On site energy resources
Transportation
Waste and pollution Water conservation
Innovative reduction water
technologies
Water use
Water efficient landscape/
external
Energy efficiency Building envelope performance
Renewable energy
Natural lighting/lighting
Energy-efficient heating/cooling
system
Mechanical systems
Green house gases emission
Machines/appliances
Material Local/regional materials
Renewable material
Recycle material
Resource reuse
Environmental impact of
materials
Economical Cost and economic Material and construction
Waste management
Fig. 2. Steps in the AHP method. Water efficiency
Site
Energy efficiency
Table 3 Social Accessibilities Accessibility to urban amenities
Preference scale for pairwise comparisons. Accessibility to public
transportation
Linguistic term Numerical value Externalities Available Services
Equally preferred 1 Social Cost Benefit Analysis
Equally to moderately preferred 2 Local Employment Opportunities
Moderately preferred 3 Indoor environment Natural ventilation efficiency
Moderately to strong preferred 4 quality Acoustic comfort
Strongly preferred 5 Lighting comfort
Strongly to very strongly preferred 6 Thermal comfort
Very strongly preferred 7 Acoustic and noise control
Very strongly to extremely preferred 8 Indoor air quality performance
Extremely preferred 9 Occupant health and safety
Occupants satisfaction Access to View
Privacy
Human Interactions
conditions. By encoding the human experience in a series of Interior Qualities
decision-making rules, fuzzy logic is able to produce a smooth out-
put surface for all input combinations without an explicit model of
the process involved [51].
Fuzzy inference is the process of formulating the mapping from discerned. There are many types of fuzzy inference systems. The
a given input to an output using fuzzy logic. The mapping then pro- two most popular types of FIS are the Mamdani-type [42] and
vides a basis from which decisions can be made, or patterns Sugeno-type [62].

Fig. 3. Steps in a fuzzy inference system.

 M. Nilashi et al. / Knowledge-Based Systems 86 (2015) 194–209 201

Table 5 Input fuzzification involves conversion of the numerical input

SE2 indicators, their parameters and weights using AHP. values (called crisp values) into linguistic variables. Linguistic vari-
Dimension Indicators Parameters Collective ables are labels such as ‘Low’ or ‘High’ which relate more to the
judgment way that a human would describe the value. Each input has its
(Weight) own set of MFs which define how input values are mapped to
Environmental Site selection Site design 0.30 the linguistic variables. As opposed to traditional crisp sets, fuzzy
Land use 0.12 sets allow data points to be partial members of a subset and mem-
Landform 0.15
Microclimate 0.10
bers of more than one subset. The degree to which an input value
Onsite energy 0.15 belongs to a particular subset is referred to as the degree of asso-
resources ciation and takes a value anywhere from zero (no association) to
Transportation 0.18 one (complete association). Once the inputs are fuzzified into lin-
Waste and Water conservation 0.16
guistic variables, they can be used for evaluating rules without
pollution Innovative reduction 0.39
water technologies explicitly referring to the underlying numerical values. The basic
Water use 0.18 structure of a rule is given as decision-making statement. The pur-
Water efficient 0.27 pose of the defuzzification process is to combine the outputs from
landscape/external each rule into a single crisp output value for the whole network. If
Energy Building envelope 0.18
the outputs of the rules are membership functions, then the firing
efficiency performance
Renewable energy 0.21
Natural 0.17
lighting/lighting
Energy-efficient 0.16 Table 6
heating/cooling Important parameters for SE2 selected by AHP method.
system
Dimension Indicators Parameters Collective
Mechanical systems 0.08
judgment
Greenhouse gases 0.11
(Weight)
emission
Machines/appliances 0.09 Environmental Site selection Site design 0.30
Material Local/regional 0.06 Landform 0.15
materials Onsite energy 0.15
Renewable material 0.19 resources
Recycle material 0.16 Transportation 0.18
Resource reuse 0.25 Waste and Water conservation 0.16
Environmental impact 0.34 pollution Innovative reduction 0.39
of materials water technologies
Water use 0.18
Economical Cost and Material and 0.11
Water efficient 0.27
economic construction
landscape/external
Waste management 0.22
Energy Building envelope 0.18
Water efficiency 0.18
efficiency performance
Site 0.26
Renewable energy 0.21
Energy efficiency 0.23
Natural 0.17
Social Accessibilities Accessibility to urban 0.44 lighting/lighting
amenities Energy-efficient 0.16
Accessibility to public 0.56 heating/cooling
transportation system
Externalities Available Services 0.25 Material Renewable material 0.19
Social Cost Benefit 0.34 Recycle material 0.16
Analysis Resource reuse 0.25
Local Employment 0.41 Environmental impact 0.34
Opportunities of materials
Indoor Natural ventilation 0.16
Economical Cost and Waste management 0.22
environment efficiency
economic Water efficiency 0.18
quality Acoustic comfort 0.05
Site 0.26
Lighting comfort 0.17
Energy efficiency 0.23
Thermal comfort 0.07
Acoustic and noise 0.16 Social Accessibilities Accessibility to urban 0.44
control amenities
Indoor air quality 0.18 Accessibility to public 0.56
performance transportation
Occupant health and 0.21 Externalities Available Services 0.25
safety Social Cost Benefit 0.34
Occupants Access to View 0.29 Analysis
satisfaction Privacy 0.36 Local Employment 0.41
Human Interactions 0.12 Opportunities
Interior Qualities 0.23 Indoor Natural ventilation 0.16
environment efficiency
quality Lighting comfort 0.17
The general process of the fuzzy inference and its schematic Acoustic and noise 0.16
diagram is shown in Fig. 3. The ’rule base’ contains a number of control
fuzzy decision-making, fuzzy rules and the ‘database’ defines the Indoor air quality 0.18
performance
MFs of the fuzzy sets used in the fuzzy rules [49]. Usually, the rule Occupant health and 0.21
base and the database are jointly referred to as the ‘Knowledge safety
base’. The ‘decision-making unit’ performs the inference operations Occupants Access to View 0.29
on the rules and two interfaces perform fuzzification and defuzzi- satisfaction Privacy 0.36
Interior Qualities 0.23
fication, respectively.
202 M. Nilashi et al. / Knowledge-Based Systems 86 (2015) 194–209

strength of the rule is used to calculate a truncated area under- influence on sustainable development practices. The surveyed
neath the membership function. The areas from each rule are then experts made their judgments based on their professional experi-
combined and the centroid is taken as the output value. For faster ence and the information provided about the green building’s
processing, the center of area calculation can be modified from a characteristics.
true centroid calculation to simpler approximations. Suppose that if there are m complete questionnaires (m = 12)
and n indicators which to be weighted by AHP, the expert e could
provide pairwise comparison matrix as:
4. Empirical study
0 1
ae11 ae12 . . . ; aen
4.1. AHP B ae ae22 e C
. . . ; a2n C
B 21
Ae ¼ B
B .. ..
C
.. C
As we discussed earlier, in this study we applied AHP and fuzzy @ . . . A
logic in developing a model for performance assessment of a green aen1 aen2 . . . ; aenn
building. AHP as a multi-criteria decision making technique which
is used for evaluating and weighting the components of a model.
where Ae is the pairwise comparison matrix completed by the
Hence, in the first step we provide a list of dimensions, indicators
expert e (e = 1 . . . m), and aeij indicates relative importance between
and parameters which all indicators and relevant parameters have
indicator i (i = 1, 2, . . . , n) and indicator j (j = 1, 2, . . . , n) based on the
been identified from the literature as reviewed in Section 2. As the
judgment of expert e and comparison value aeji ¼ a1e .
aim of this study was to develop a new rating system for green ij

building performance assessment, the use of in-depth interview
with the building expert along with conducting a questionnaire
survey as a convenient way to collect data were necessary to
enable the study to get a clear picture of what is actual phenomena
in the real situation. Hence, we contacted individual experts sepa-
rately to interview and obtain a completed questionnaire. For each
expert, a questionnaire was developed which included comparison
matrices along with explanations about components. All 12
experts that participated in the survey had at least 5 years’ experi-
ence in the building field, and had more than 10 years’ professional
experience. They were selected according to their role and
The pairwise comparison matrices are developed by the experts
by using the scale given in Table 3 which are preference scale for
pairwise comparisons recommend by Saaty. As an example, Site
design and Transportation were compared under Site selection
using the question ‘‘How important is Site design when it is com-
pared with Transportation?’’ and the given answer by expert was
‘‘Moderately Preferred’’, instead of the numerical value. Then we
replaced corresponding numerical value of judgment in the rele-
vant cell of comparison matrices (Ae). Following this procedure,
the matrix of each expert’s judgment was established and weight
for each parameter was calculated.

1 1

0.8 Low Modeate High 0.8 Modeate High

Low

0.6 0.6

0.4 0.4

0.2 0.2

0
0 1 2 3 4 5 6 7 8 0
0 0.5 1 1.5 2 2.5 3 3.5 4
Input Variable "Site Selection"
Input Variable "Accessibilities"

1 1

0.8 Low Modeate High 0.8 Low Moderate High

0.6 0.6

0.4 0.4

0.2 0.2

0 0
0 2 4 6 8 10 0 1 2 3 4 5 6 7 8
Input Variable "Indoor Environment Quality" Input Variable "Environmental"

1 1

0.8 Low Modeate High 0.8 Low Modeate High

0.6 0.6

0.4 0.4

0.2 0.2

0 0
0 0.5 1 1.5 2 0 1 2 3 4 5 6 7 8
Input Variable "Economical" Input Variable "Social"

Fig. 4. Membership functions for input variables of fuzzy system.

 M. Nilashi et al. / Knowledge-Based Systems 86 (2015) 194–209 203

As can be seen, the list of indicators and parameters in each of the parameters in the three dimensions. To obtain a collective
dimension is presented in Table 4. In order to calculate the rank judgment, geometric mean method was used to aggregate individ-
of parameters in AHP method, after collecting the pair comparison ual judgment. The geometric mean method for n element of x1, x2,
questionnaires, Expert Choice 2000 software was used. Based on . . . , xn is presented in Eq. (1).
threshold decided by experts, the most important parameters in qffiffiffiffiffiffiffiffiffiffiffiffiffiffi
each dimension were selected. Table 5 summarizes the weights GM ¼
n
Pni¼1 xi ð1Þ

1 1

0.8 Vlow Low Modeate High Vhigh 0.8 Vlow Low Modeate High Vhigh

0.6 0.6

0.4 0.4

0.2 0.2

0 0
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
Output Variable "Economical" Output Variable "Social"

1 1

0.8 Vlow Low Modeate High Vhigh 0.8 Vlow Low Modeate High Vhigh

0.6 0.6

0.4 0.4

0.2 0.2

0 0
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
Output Variable "Environmental" Output Variable "Performance"

Fig. 5. Membership functions for output variables of fuzzy system.

Fig. 6. Fuzzy models structure for green building performance assessment.

204 M. Nilashi et al. / Knowledge-Based Systems 86 (2015) 194–209

From the weights in Table 5, the most important parameters are From the relationship, we can define that the performance level
selected with threshold value 0.15. Hence, Site design, Landform, of a green building is a function of SE2.
Onsite energy resources and Transportation with weights 0.30, As totally 33 parameters and 9 indicators have been considered
0.15, 0.15 and 0.18 are selected for the Site selection indicator. for constructing the fuzzy logic assessment model, the Likert-based
For the Waste and pollution, all parameters are selected as their questionnaire has been designed based on these parameters and
weights are greater than the threshold value. For the Energy effi- indicators to collect the data from the second group (totally 120
ciency, except Mechanical systems and Greenhouse gases emission respondents) of respondents. The collected data from this group
and Machines/appliances with weights 0.08, 0.11 and 0.09, other is used for forming the fuzzy rules (discovering knowledge) to be
parameters are selected. For this indicator, Renewable energy is used in the FIS of fuzzy logic-based assessment model.
the most important parameter with weight 0.21. For the Material The fuzzy model based on Mamdani algorithm is implemented
indicator, Local/regional materials is rejected to be considered for on fuzzy logic toolbox of MATLAB software package. The proposed
the next evaluation as in relation to the other parameters in this system includes two main tiers. In the first tier, the system
group, it has the lowest weight (0.06) based on experts’ judgment. assesses the SE2 levels of a green building. Then, in the second tier,
In the Economical dimension and for the Cost and economic indi- the performance level is then evaluated based on the obtained
cator, Waste management, Water efficiency, Site and Energy effi- levels of SE2. To be able to develop the fuzzy logic-based perfor-
ciency were important with weights 0.22, 0.18, 0.26 and 0.23. In mance assessment model, the appropriate MFs should be defined
Social dimension, all parameters for the Accessibilities and for the inputs and outputs (fuzzifying the variables) of model in
Externalities indicators were selected, however, for Indoor envi- Fuzzy Inferences System (FIS). In this research, all input variables
ronment quality the Acoustic comfort and Thermal comfort indica- in the FIS model uses linguistic terms of Low, Moderate and High
tors obtained 0.05 and 0.07 which cannot be considered based on with their MFs modeled as Gaussian MFs (see Fig. 4). In addition,
the threshold value. Finally, for the Occupants Satisfaction indica- for the outputs of model, the Triangular MFs are considered as pre-
tor, Access to View, Privacy, Interior Qualities are selected with sented in (see Fig. 5). They are defined with the linguistic variables
weights 0.29, 0.36 and 0.23, respectively. For this indicator, as Vlow, Low, Moderate, High and Vhigh. The structure of the fuzzy
Privacy has been considered the most important factor in relation models for green building performance assessment is shown in
to the Access to View and Interior Qualities. In Table 6, the impor- Fig. 6. Tables 7 and 8 present membership functions for input (indi-
tant parameters for SE2 selected by AHP method are presented. cators) and output variables, respectively.
Based on the MFs for inputs and outputs, the linguistic variables
4.2. Fuzzy logic-based model for inputs and outputs have been considered in three and five
levels, respectively. Using the defined MFs, the fuzzy model can
This study is based on the rationale that actual performance generate the real level of green building performance based on dis-
level of a green building is based on three factors, namely: what covered fuzzy rules in the FIS system. It should be noted that
is the Social performance level (S) of the given green building, what selecting the MF types and its ranges for the inputs and outputs
is the Environmental performance level (E) of the given green play an important role in the performance assessment. Thus, based
building and what is the Economical performance level (C) of a on our experience and collected data from the experts, we selected
given green building? Therefore, we propose a fuzzy logic-based different ranges and types for MFs. In addition, the Gaussian type
assessment model for detecting the performance of a given green of MFs are selected for inputs because of being the most natural
building according to the following relationship: [43], smooth and nonzero at all points [67].
After defuzzification process of input and output variables, in
Lperformance ðGreen BuildingsÞ ¼ f ðS; E; CÞ ð2Þ
the next step, the construction of the fuzzy IF-THEN rules was
formed. The identified fuzzy rules represent the fuzzy relations
between input and output variables. Based on experts’ knowledge,
Table 7 the rule base of the fuzzy model was constructed. In this study, the
Membership functions for input variables (indicators).
fuzzy rules used for performance assessment have been discovered
Main factor Indicator Number of Linguistic terms for fuzzy system from the experts’ knowledge and for each tier different fuzzy rules
number number parameter have been identified. In Table 9, the number of fuzzy rules is pre-
Low Moderate High
sented. It can be seen in Table 9, because the model in the first tier
1 1 4 [1.3590] [1.3594] [1.3598]
2 4 [1.3590] [1.3594] [1.3598] has 4 inputs for Environmental, 1 input for Economical and 4
3 4 [1.3590] [1.3594] [1.3598] inputs for Social; thus, 165 fuzzy rules are included in the FIS. In
4 4 [1.3590] [1.3594] [1.3598] addition, in the second tier, as the model has 3 inputs; thus, 27
2 1 4 [1.3590] [1.3594] [1.3598] fuzzy rules are incorporated in the FIS. Therefore, totally 192 fuzzy
3 1 2 [0.67930] [0.67932] [0.67934] rule are considered for green building performance assessment in
2 3 [1.6 0 1.6] [1.0193] [1.0196] FIS. It should be noted that the number of fuzzy rules can be con-
3 5 [1.6990] [1.6995] [1.69910] trolled by increasing or decreasing the number of MFs.
4 3 [1.6 0 1.6] [1.0193] [1.0196] Accordingly, a formation of fuzzy rules for performance

Table 8
Membership functions for output variables.

Output Linguistic terms for fuzzy system

Vlow Low Moderate High Vhigh
Environmental [0.2 0 0.2] [0 0.2 0.4] [0.3 0.5 0.7] [0.6 0.81] [0.8 1 1.2]
Social [0.2 0 0.2] [0 0.2 0.4] [0.3 0.5 0.7] [0.6 0.81] [0.8 1 1.2]
Economical [0.2 0 0.2] [0 0.2 0.4] [0.3 0.5 0.7] [0.6 0.81] [0.8 1 1.2]
Performance [0.2 0 0.2] [0 0.2 0.4] [0.3 0.5 0.7] [0.6 0.81] [0.8 1 1.2]
 M. Nilashi et al. / Knowledge-Based Systems 86 (2015) 194–209 205

Table 9 model for inference mechanism that Aggregation method between

Number of fuzzy rules for FIS model of assessment. rules is maximum to combine output fuzzy set, so Fuzzification
Inputs Output Number of fuzzy rules method here is max–min and Defuzzification method is centroid.
Site selection Environmental 3  3  3  3 = 81 For model simulation, the GUI tools of Fuzzy Logic Toolbox library
Waste and pollution was embedded into the Simulink. Fuzzy Logic Toolbox library
Energy efficiency which contains the Fuzzy Logic Controller and Fuzzy Logic
Material Controller with Rule Viewer blocks was used for employing the
Cost and economic Economical 31=3 FIS models (containing 192 fuzzy rules) of performance assessment
Accessibilities 3  3  3  3 = 81 system. Simulink block diagram for green building performance
Externalities
Indoor Environment
assessment system is shown in Fig. 7. As can be seen, the
Quality Simulink block diagram contains four Fuzzy Logic Controller with
Occupants Satisfaction Rule Viewer blocks, nine constant blocks, three multiplexer blocks
Environmental Performance 3  3  3 = 27 and a display window for output.
Economical Combining both the input MFs and the output MFs with the
Social rules presented in Table 10, 2D curves and 3D plots can be
Total number of fuzzy rules: obtained to give a snapshot relationship between the inputs and
192 output. Illustrating the interdependency between inputs and out-
put is helpful in revealing the level of green building performance.
Fig. 8 illustrates the interdependency of performance and three
assessment is provided in Table 10 for 3 main inputs (SE2) and one main dimensions through the curves and control surfaces obtained
output (performance). from the fuzzy rules and discovered from the collected data from
The fuzzy model is implemented in fuzzy logic toolbox of experts. The level of performance can be depicted as a continuous
MATLAB software package. This system depends on Mamdani function of its input parameters as Environmental, Economical, and

Table 10
Fuzzy rules formation for performance assessment.

Rule # IF Environmental AND Social AND Economical THEN Performance level

Linguistic value Linguistic value Linguistic value Linguistic Value
1 High Low Low Moderate
2 High Moderate High Very high
3 Low Moderate High Low
4 Low Low High Very low
. . . . .
. . . . .
. . . . .
24 Low Low Low Very low
25 Low High Moderate Moderate
26 High High Moderate Very high
27 High Moderate Low Moderate

Fig. 7. Simulink block diagram for green building performance assessment system.
206 M. Nilashi et al. / Knowledge-Based Systems 86 (2015) 194–209
Social. The curves and control surfaces depict the variation of per-
formance based on identified rules. The control surfaces also depict
the interdependency of performance on ‘‘Environmental–Econom
ical’’ and ‘‘Environmental–Social’’ and ‘‘Social–Economical’’.
These surface plots show exactly the performance level on any
two dimensions of a green building. They also indicate that which
dimension can be important for the performance level. It should be
noted that from the surface plots and according to the discovered
fuzzy rules, the level of performance is more dependent on
Environment dimension. The level of performance level is obtained
in its maximum level about 0.911. However, for two other dimen-
sions Economical and Social, the performance level is obtained
about 0.513 and 0.603, respectively. This shows that the Social
dimension for performance level is also more important than
Economical factor. Here, we can say that assessing the performance
level of a given green building can be modeled by FIS based on
experts knowledge which has been formed as fuzzy rules.
In order to composite fuzzy relations, the max–min composi-
tion method, most popular applied technique [55], is selected.
Fig. 8. Interdependency between inputs and output of fuzzy system.
The centroid of area (COA) method is the most widely adopted
defuzzification method [25]. It is essentially the weighted average
of the outcome fuzzy set.
Using COA, the Rule Viewer module of the FIS presents an inter-
pretation of the entire fuzzy inference process regarding the per-
formance of a green building. The fuzzy rule viewer of the
established model is shown in Fig. 9 for indicating the performance
of green building over the change in values of all three inputs. From
the fuzzy rule viewer above, for example, we can see when the
input parameters of Environmental is at 6.04, Economical at 2,
and Social at 3.48, an output of performance level at 0.861 is
obtained.
It should be noted that the heart of the proposed performance
assessment model using FIS is the knowledge-base that forms
the Inference Engine. In this case, the knowledge of the human
expert is translated to decision-making rules used in the
Inference Engine. It can be argued that the level of expertise from
one human expert in green building may differ from another and
this form the different fuzzy rules but as long as they are fixed in
 M. Nilashi et al. / Knowledge-Based Systems 86 (2015) 194–209
Fig. 9. Fuzzy rule viewer for inputs and output variables of fuzzy model.
the Inference Engine, the outputs of FIS will always be consistent.
In Fig. 9, it can be also seen that the FIS works by converting qual-
itative indicators of the green building performance assessment to
numerical values that make it possible to quantify the perfor-
mance’ level. Apart from that, it is also possible to store these
numerical values in a database for the new evaluations.
5. Conclusions and further study
In this research, an effort has been made to develop an expert
system to assess the performance level of green buildings using
AHP and fuzzy logic approaches. The assessment criteria have been
selected from the literature based on the three main dimensions of
assessment, Social, Economical and Environmental denoted as SE2
in this research work. The data for this research has been collected
from the experts in the field via pair-wise and Likert-based ques-
tionnaires. To select the most important factors in each dimension,
we applied AHP prior to the fuzzy logic approach as fuzzy-AHP.
Then, based on selected factors, an efficient performance assess-
ment system based on fuzzy logic has been developed along with
its FIS to assess the level of performance. All input variables in
the FIS model used linguistic terms modeled as Gaussian MFs
and for the outputs of model, we considered the Triangular MFs.
With defining these MFs for all inputs and outputs of fuzzy system,
192 fuzzy rules were discovered to be used in the FIS. The results of
proposed expert system showed the ability of fuzzy logic in evalu-
ating the performance level of green building. In addition, the eval-
uation results also demonstrated that the Environmental
dimension is more important for performance level in relation to
the Social and Economical dimensions based on the experts
knowledge.
In this study, some implications and limitations exist which
need to be focused and scrutinized in further studies. First, there
were a small number of experts in fulfilling the survey for this
study. As sophisticated analysis is derived by the large sample size
207
of the respondents, there is a call for future study conducting a rig-
orous study to examine the evaluation of Green Buildings from a
large number of respondents; this also lead to generalize findings
of the prospective study. Hence, from this limitation, we feel that
collecting the data from the multiple sources can improve the
scope of generalizability of the proposed fuzzy model. Second, in
this research, the output of the FIS was restricted by the design
and number of MFs and rules of the knowledge-base in the
Inference Engine. In this particular design, Triangular and
Gaussian MFs were used to represent the linguistic terms to assess
the performance level. Thus, different MFs such as Trapezoidal or
Sigmoid may yield different outputs and therefore they should be
investigated, and we propose inclusion of this aspect in future
research.
Acknowledgments
This work is supported by the Fundamental Research Grant
Scheme (FRGS), granted financial resource from the Ministry of
Education (MOE), Malaysia (Ref No.: PY/2014/04114, Vot No.:
4F680). Appreciation also goes to the anonymous reviewers whose
comments helped us to improve the manuscript.
References
[1] H. Ahmadi, M. Nilashi, O. Ibrahim, Prioritizing critical factors to successful
adoption of total hospital information system, J. Soft Comput. Decis. Support
Syst. 2 (4) (2015) 6–16.
[2] H. Ahmadi, M. Nilashi, O. Ibrahim, Organizational decision to adopt hospital
information system: an empirical investigation in the case of Malaysian public
hospitals, Int. J. Med. Inform. 84 (3) (2015) 166–188.
[3] H. Ahmadi, M.S. Rad, M. Nazari, M. Nilashi, O. Ibrahim, Evaluating the factors
affecting the implementation of hospital information system (HIS) using AHP
method, Life Sci. J. 11 (3) (2014).
[4] H.H. Ali, S.F. Al Nsairat, Developing a green building assessment tool for
developing countries–Case of Jordan, Build. Environ. 44 (5) (2009) 1053–1064.
208 M. Nilashi et al. / Knowledge-Based Systems 86 (2015) 194–209
[5] S.H. Alyami, Y. Rezgui, A. Kwan, Developing sustainable building assessment
scheme for Saudi Arabia: Delphi consultation approach, Renew. Sust. Energy
Rev. 27 (2013) 43–54.
[6] G. Assefa, M. Glaumann, T. Malmqvist, B. Kindembe, M. Hult, U. Myhr, O.
Eriksson, Environmental assessment of real estates–where natural and social
sciences meet: the case of ecoeffect, in: Proceedings of the 2005 World
Sustainable Building Conference (SB05Tokyo), 2005, pp. 27–29.
[7] BREEAM. Available at: <http://www.breeam.org/page.jsp?id=66>, 2010
(accessed December 2011).
[8] V. Behbood, J. Lu, G. Zhang, Adaptive Inference-based learning and rule
generation algorithms in fuzzy neural network for failure prediction, in: 2010
International Conference on Intelligent Systems and Knowledge Engineering
(ISKE), IEEE, 2010, pp. 33–38.
[9] U. Berardi, Sustainability assessment in the construction sector: rating systems
and rated buildings, Sust. Dev. 20 (6) (2012) 411–424.
[10] C. Carlsson, R. Fullér, Fuzzy multiple criteria decision making: recent
developments, Fuzzy Sets Syst. 78 (2) (1996) 139–153.
[11] S.R. Chandratilake, W.P.S. Dias, Sustainability rating systems for buildings:
comparisons and correlations, Energy 59 (2013) 22–28.
[12] P.L. Chang, Y.C. Chen, A fuzzy multi-criteria decision making method for
technology transfer strategy selection in biotechnology, Fuzzy Sets Syst. 63 (2)
(1994) 131–139.
[13] C.K. Chau, M.S. Tse, K.Y. Chung, A choice experiment to estimate the effect of
green experience on preferences and willingness-to-pay for green building
attributes, Build. Environ. 45 (11) (2010) 2553–2561.
[14] Z. Chen, D.J. Clements-Croome, J. Hong, H. Li, Q. Xu, A multicriteria lifespan
energy efficiency approach to intelligent building assessment, Energy Build. 38
(5) (2006) 393–409.
[15] D.J. Clements-Croome, Intelligent Buildings: Design, Management and
Operation, Thomas Telford, London, 2004.
[16] R.J. Cole, Emerging trends in building environmental assessment methods,
Build. Res. Inform. 26 (1) (1998) 3–16.
[17] DGNB German Sustainable Building Certificate. Structure – Application –
Criteria. German Sustainable Building Council, 2009.
[18] A. Forsberg, F. Malmborg, Tools for environmental assessment of the built
environment, Build. Environ. 39 (2) (2004) 223–228.
[19] K.M. Fowler, E.M. Rauch, Sustainable building rating systems summary. Pacific
Northwest National Laboratory; 2006 July.
[20] GBI Assessment Criteria for Non-Residential Exisiting Building (NREB), first ed.
January 2011, version 1.1. <http://www.greenbuildingindex.org/> (accessed
on 10.8.11).
[21] Z. Gou, S.S.Y. Lau, Contextualizing green building rating systems: case study of
Hong Kong, Habitat Int. 44 (2014) 282–289.
[22] T. Griffin, T. DeLacey, R. Harris, T. Griffin, P. Williams, Green Globe:
Sustainability Accreditation for Tourism, 2002.
[23] A.Z. Hamedani, F. Huber, A comparative study of DGNB, LEED and BREEAM
certificate systems in urban sustainability, Sust. City VII: Urban Regen. Sust.
1121 (2012).
[24] J. Heerwagen, Green buildings, organizational success and occupant
productivity, Build. Res. Inform. 28 (5/6) (2000) 353–367.
[25] H. Hellendoorn, C. Thomas, Defuzzification in fuzzy controllers, J. Intell. Fuzzy
Syst. 1 (2) (1993) 109–123.
[26] H.M. Hsu, C.T. Chen, Aggregation of fuzzy opinions under group decision
making, Fuzzy Sets Syst. 79 (3) (1996) 279–285.
[27] O. Ibrahim, M. Nilashi, K. Bagherifard, N. Hashem, N. Janahmadi, M. Barisami,
Application of AHP and KMeans clustering for ranking and classifying
customer trust in M-commerce, Aust. J. Basic Appl. Sci. 5 (12) (2011).
[28] M. Kabak, M. Dağdeviren, Prioritization of renewable energy sources for
Turkey by using a hybrid MCDM methodology, Energy Convers. Manage. 79
(2014) 25–33.
[29] J.T. Kim, Sustainable and healthy buildings, Energy Build. 46 (2012) 1–2.
[30] G. Kim, H.S. Lim, T.S. Lim, L. Schaefer, J.T. Kim, Comparative advantage of an
exterior shading device in thermal performance for residential buildings,
Energy Build. 46 (2012) 105–111.
[31] C. Koo, T. Hong, M. Lee, H.S. Park, Development of a new energy efficiency
rating system for existing residential buildings, Energy Policy 68 (2014) 218–
231.
[32] H.W. Kua, S.E. Lee, Demonstration intelligent building – a methodology for the
promotion of total sustainability in the built environment, Build. Environ. 37
(3) (2002) 231–240.
[33] A.C. Kutlu, M. Ekmekçioğlu, Fuzzy failure modes and effects analysis
by using fuzzy TOPSIS-based fuzzy AHP, Exp. Syst. Appl. 39 (1) (2012)
61–67.
[34] P. Lam, E. Chan, C. Poon, C. Chau, K. Chun, Factors affecting the implementation
of green specifications in construction, J. Environ. Manage. 91 (3) (2010)
654–661.
[35] M. Laumanns, E. Zitzler, L. Thiele, Multiple criteria decision support by
evolutionary computation, in: Sustainability in the Information Society, 15th
International Symposium Informatics for Environmental Protection, Zurich,
2001.
[36] T. Litman, D. Burwell, Issues in sustainable transportation, Int. J. Global
Environ. Issues 6 (4) (2006) 331–347.
[37] E. Loh, T. Crosbie, N. Dawood, J. Dean, A framework and decision support
system to increase building life cycle energy performance, J. Inform. Technol.
Constr. (2010). ISSN 1874-4753.
[38] P.E. Love, M. Niedzweicki, P.A. Bullen, D.J. Edwards, Achieving the green
building council of Australia’s world leadership rating in an office building in
Perth, J. Constr. Eng. Manage. 138 (5) (2011) 652–660.
[39] J. Lu, J. Ma, G. Zhang, Y. Zhu, X. Zeng, L. Koehl, Theme-based comprehensive
evaluation in new product development using fuzzy hierarchical criteria
group decision-making method, IEEE Trans. Ind. Electron. 58 (6) (2011) 2236–
2246.
[40] J. Lu, G. Zhang, D. Ruan, F. Wu, Multi-objective Group Decision Making.
Methods, Software, and Applications with Fuzzy set Techniques, Imperial
College Press, 2007.
[41] J. Ma, J. Lu, G. Zhang, Decider: a fuzzy multi-criteria group decision support
system, Knowl.-Based Syst. 23 (1) (2010) 23–31.
[42] E.H. Mamdani, Application of fuzzy algorithms for control of simple dynamic
plant, in: Proceedings of the Institution of Electrical Engineers, vol. 121, No. 12,
IET Digital Library, 1974, pp. 1585–1588.
[43] A.S. Markowski, M.S. Mannan, Fuzzy risk matrix, J. Hazard. Mater. 159 (1)
(2008) 152–157.
[44] M. Marzouk, A. Nouh, M. El-Said, Developing green bridge rating system using
Simos’ procedure, HBRC J. 10 (2) (2014) 176–182.
[45] R. Mateus, L. Braganca, Sustainability assessment and rating of buildings:
developing the methodology SBTool, Build. Environ. 46 (10) (2011) 1962–
1971.
[46] M. Medineckiene, E.K. Zavadskas, F. Björk, Z. Turskis, Multi-criteria decision-
making system for sustainable building assessment/certification, Arch. Civil
Mech. Eng. 15 (1) (2015) 11–18.
[47] A. Meier, T. Olofsson, R. Lamberts, What is an, 2002.
[48] B. Michael, M. Peter, S. Michael, Green Building: Guidebook for
Sustainable Architecture, Springer, Heidelberg, Dordrecht, London, New York,
2007.
[49] M. Naderpour, J. Lu, A fuzzy dual expert system for managing situation
awareness in a safety supervisory system, in: 2012 IEEE International
Conference on Fuzzy Systems (FUZZ-IEEE), IEEE, 2012, pp. 1–7.
[50] B.K. Nguyen, H. Altan, Comparative review of five sustainable rating systems,
Proc. Eng. 21 (2011) 376–386.
[51] H.T. Nguyen, E.A. Walker, A First Course in Fuzzy Logic, CRC Press, 2005.
[52] M. Nilashi, N. Janahmadi, Assessing and prioritizing affecting factors in e-
learning websites using AHP method and fuzzy approach, Inform. Knowl.
Manage. 2 (1) (2012).
[53] L. Pérez-Lombard, J. Ortiz, C. Pout, A review on buildings energy consumption
information, Energy Build. 40 (3) (2008) 394–398.
[54] J.H. Purba, J. Lu, G. Zhang, W. Pedrycz, A fuzzy reliability assessment of basic
events of fault trees through qualitative data processing, Fuzzy Sets Syst. 243
(2014) 50–69.
[55] T.J. Ross, Fuzzy Logic with Engineering Applications, John Wiley & Sons,
2004.
[56] O. Saadatian, L.C. Haw, S.B. Mat, K. Sopian, Perspective of sustainable
development in Malaysia, Int. J. Energy Environ. 2 (6) (2012) 260–267.
[57] T.L. Saaty, Decision making with the analytic hierarchy process, Int. J. Serv. Sci.
1 (1) (2008) 83–98.
[58] S.R. Sahamir, R. Zakaria, Green assessment criteria for public hospital building
development in Malaysia, Proc. Environ. Sci. 20 (2014) 106–115.
[59] M. Salahshour, H.M. Dahlan, N.A. Iahad, M. Nilashi, O. Ibrahim, Using a multi-
criteria decision making approach for assessing the factors affecting social
network sites intention to use, J. Soft Comput. Decis. Support Syst. 2 (3) (2015)
20–28.
[60] J. Šelih, A. Kne, A. Srdić, M. Žura, Multiple-criteria decision support system in
highway infrastructure management, Transport 23 (4) (2008) 299–305.
[61] A. Sinha, R. Gupta, A. Kutnar, Sustainable development and green buildings,
Faculty Research Publication of College of Forestry & Oregon Forest Research
Laboratory, 2013. <http://hdl.handle.net/1957/40374>.
[62] T. Takagi, M. Sugeno, Fuzzy identification of systems and its applications to
modeling and control, IEEE Trans. Syst. Man Cybernet. 1 (1985) 116–132.
[63] M.S. Todorovic, J.T. Kim, Beyond the science and art of the healthy buildings
daylighting dynamic control’s performance prediction and validation, Energy
Build. 46 (2012) 159–166.
[64] P. Vihakapirom, R. Li, A framework for distributed group multi-criteria
decision support systems, The Ninth Australian World Wide Web
Conference (Andrew Treloar 05 July 2003 to 09 July 2003), vol. 158,
Southern Cross University, 2003, pp. 1–8.
[65] Y.J. Wang, H.S. Lee, K. Lin, Fuzzy TOPSIS for multi-criteria decision making, Int.
Math. J. 3 (2003) 367–379.
[66] J. Wong, H. Li, Development of a conceptual model for the selection of
intelligent building systems, Build. Environ. 41 (8) (2006) 1106–1123.
[67] M. Xie, Fundamentals of Robotics: Linking Perception to Action, World
Scientific Publishing Co Ltd, London, 2003.
[68] J. Xu, Y. Wang, Z. Tao, Rough approximation based strategy model between a
green building developer and a contractor under a fuzzy environment, Knowl.-
Based Syst. 46 (2013) 54–68.
[69] E.J. Yanarella, R.S. Levine, R.W. Lancaster, Research and solutions: ‘‘Green’’ vs.
sustainability: from semantics to enlightenment, Sust.: J. Rec. 2 (5) (2009)
296–302.
[70] Yong Han Ahn, The Development of Models to Identify Relationships Between
First Costs of Green Building Strategies and Technologies and Life Cycle Costs
for Public Green Facilities. PhD thesis: Virginia Polytechnic Institute and State
University, 2010.
 M. Nilashi et al. / Knowledge-Based Systems 86 (2015) 194–209
[71] S.W. Yoon, D.K. Lee, The development of the evaluation model of climate
changes and air pollution for sustainability of cities in Korea, Landscape Urban
Plan. 63 (3) (2003) 145–160.
[72] W. Yu, B. Li, X. Yang, Q. Wang, A development of a rating method and
weighting system for green store buildings in China, Renew. Energy 73 (2015)
123–129.
[73] J. Yudelson, The Green Building Revolution, Island Press, Washington, DC,
2007.
[74] L.A. Zadeh, Outline of a new approach to the analysis of complex systems and
decision processes, IEEE Trans. Syst. Man Cybernet. 1 (1973) 28–44.
209
[75] R. Zakaria, L.Z. Kwong, M. Nilashi, A. Majid, M. Zaimi, O. Ibrahim, R.M. Zin,
Ethical behaviors in E-tendering process for construction project in Malaysia, J.
Theor. Appl. Inform. Technol. 70 (1) (2014).
[76] J. Zhang, J. Lu, G. Zhang, A hybrid knowledge-based prediction method for
avian influenza early warning, in: IEEE International Conference on Systems,
Man and Cybernetics, 2009, SMC 2009, IEEE, 2009, pp. 617–622.
[77] R. Zhang, J. Lu, G. Zhang, A knowledge-based multi-role decision support
system for ore blending cost optimization of blast furnaces, Euro. J. Oper. Res.
215 (1) (2011) 194–203.