International Journal on Semantic Web and Information Systems

Volume 20 • Issue 1 • January-December 2024

Group Consensus-Driven Energy

Consumption Assessment Using
Social Network Analysis and
Fuzzy Information Fusion
Bingjie Wang Mohammed J. F. Alenazi
Shanxi University, China King Saud University, Saudi Arabia
Chao Zhang Salman A. AlQahtani
https://orcid.org/0000-0001-6248-9962 King Saud University, Saudi Arabia
Shanxi University, China
Sendhil Kumar K. S.
Arun Kumar Sangaiah https://orcid.org/0000-0001-6669-9334
National Yunlin University of Science and Technology, VIT University, India
Taiwan & Sunway University, Malaysia

ABSTRACT

With industrial development deepening, the share of industrial energy in overall consumption
has notably risen. To assess industrial energy consumption accurately, this paper proposes a method
employing interval type-2 fuzzy sets (IT2FSs) to represent assessment information effectively.
Additionally, it analyzes decision-makers (DMs) as a social network to alleviate individual biases.
IT2FSs are chosen to handle uncertainties in assessing industrial energy consumption. Addressing
biases in DMs' opinions, a group consensus model aids the consensus reaching process (CRP).
Industrial energy consumption is assessed using the MULTIMOORA method, yielding three results.
These are fused via D-S evidence theory (DSET) to obtain the final assessment. Finally, the model's
effectiveness is verified with a case study on energy consumption in the steel industry. In conclusion,
this paper not only deepens the understanding of uncertainties in the energy consumption assessment
process, but also provides a robust tool for various industries to optimize energy use and economic
outcomes.

KEYWORDS
Consumption Assessment, Interval Type-2 Fuzzy Set, Group Consensus, Social Network Analysis, Information
Fusion, MULTIMOORA, Evidence Theory, Energy Consumption Assessment

PROBLEM AND BACKGROUND

Industrial energy consumption is a cornerstone of modern economies, powering manufacturing

processes, transportation systems, and myriad other activities vital for societal functioning. However,
the sheer scale of energy usage in industrial sectors raises significant environmental, economic,
and social concerns (Zhao & Zhu, 2022). Hence, assessing industrial energy consumption becomes
imperative for ensuring energy security, mitigating environmental impact, and fostering sustainable
growth. Assessing industrial energy consumption (Sangaiah et al., 2023) offers multifaceted benefits
with tangible real-world implications. Additionally, it facilitates informed energy consumption

DOI: 10.4018/IJSWIS.352043

This article published as an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creative-
commons.org/licenses/by/4.0/) which permits unrestricted use, distribution, and production in any medium, provided the author of the
original work and original publication source are properly credited.

1
International Journal on Semantic Web and Information Systems
Volume 20 • Issue 1 • January-December 2024

Figure 1. The chart of industrial energy consumption assessment

assessment (Gupta & Panigrahi, 2022) by providing insights into energy usage patterns, allowing
policymakers, businesses, and stakeholders to identify areas of inefficiency and prioritize resource
allocation. Moreover, it enables the formulation and implementation of targeted energy efficiency
measures, thereby enhancing productivity, reducing operational costs, and bolstering competitiveness
in global markets. Furthermore, it aids environmental sustainability endeavors by pinpointing
opportunities to decrease carbon emissions, mitigate climate change, and transition to cleaner energy
sources (Rao et al., 2023). Additionally, it fosters innovation and technological advancements by
incentivizing research and development in energy-efficient technologies and processes.
The steel industry (Alrumaih & Alenazi, 2024) stands out as a prominent consumer of energy,
accounting for a significant portion of industrial energy consumption worldwide. The production of
steel involves energy-intensive processes such as iron ore smelting, coke production, and steelmaking,
which collectively contribute to substantial energy usage and carbon emissions. Consequently, the
steel industry faces unique challenges in terms of energy efficiency, resource optimization, and
environmental sustainability. Despite these challenges, the steel industry also presents significant
opportunities for energy conservation and emission reduction. By assessing its energy usage, we can
identify opportunities to reduce greenhouse gas emissions, mitigate climate change, and improve air
quality (Patil & Singh, 2024). Overall, assessing energy consumption (Bai et al., 2020) in the steel
industry is crucial for achieving environmental, economic, and social sustainability goals. Therefore,
this paper will conduct decision assessment within the context of energy consumption in the steel
industry. The schematic diagram of industrial energy consumption assessment is shown in Figure 1.
From Figure 1, it can be seen that in this energy consumption assessment, we use time points as decision
makers and dates as alternatives. The energy consumption assessment is then conducted based on
the evaluation attributes. In this paper, three classic theories are utilized. This paper chooses interval
type-2 fuzzy sets (IT2FSs) as the preferred information expression of decision makers (DMs). After
the consensus-reaching process (CRP), assessment is conducted through multiobjective optimization
on the basis of ratio analysis plus full multiplicative form (MULTIMOORA). Finally, the assessment
results are integrated through evidence theory to obtain the final outcome.

A Brief Review of IT2FSs

In response to the challenge of DMs providing vague assessment information in multiple attribute
group decision-making (MAGDM), Zadeh (1965) introduced fuzzy sets in 1965. Expanding on this

2
 International Journal on Semantic Web and Information Systems
Volume 20 • Issue 1 • January-December 2024

concept, Zadeh (1975) further developed T2FSs in 1975. This extension involves fuzzifying fuzzy sets
even more by transforming the membership functions of fuzzy sets into type-1 fuzzy sets (T1FSs),
thus effectively capturing the uncertainty inherent in the problem. IT2FSs, a specific instance of
T2FSs, are particularly adept at representing uncertainty among individuals. By utilizing IT2FSs,
not only can we capture the fuzziness in the data, but we can also account for the uncertainty about
the membership grades themselves, making it more robust in handling the vagueness present in the
data. In recent years, IT2FSs have emerged as a focal point in T2FS research, garnering attention
from numerous scholars (Jin et al., 2023; Zhang et al., 2024). Therefore, this paper selects IT2FSs
as the preference expression for DMs, thus better describing the uncertainty among DMs in energy
consumption assessment.

A Brief Review of the Group Consensus Model

This paper primarily assesses the energy consumption situation in the steel industry. Given the
complexity of this issue, the paper treats DMs as a social network for analysis. Within this social
network (Haque et al., 2023), DMs engage in discussions, exchange assessment results, and adjust
assessments that have significant biases, ultimately reaching consensus. In CRP, it is mainly divided
into two stages (Bai et al., 2024). The first stage is the measurement of group consensus, and the
second stage is the feedback mechanism. In such cases, a collaborative approach allows for the
examination of complex issues through collective input, thus maintaining the scientific and impartial
nature of the assessment process. This approach, known as the group consensus model (Zheng et
al., 2023), allows for the comprehensive assessment of multifaceted issues by leveraging the diverse
perspectives of multiple evaluators. To fully leverage the diverse perspectives and expertise of
multiple participants, thereby enhancing the comprehensiveness and accuracy of energy consumption
assessment, incorporating a group consensus model into energy consumption assessment is a favorable
choice. The advantages of the group consensus model (Zhang et al., 2024) include integrating diverse
perspectives, distributing energy consumption assessment burdens, enhancing energy consumption
assessment credibility and promoting team collaboration. In practice, the group consensus model can
be applied in various fields, including business management, project planning, policymaking, and
more to assist organizations in better addressing complex decision challenges. Therefore, this paper
proposes a group consensus model under IT2F environments. This aims to encourage cooperation
and collaboration among DMs, working together to solve problems.

A Brief Review of MULTIMOORA and D-S Evidence Theory

After CRP, the energy consumption assessment problem addressed in this paper will degrade into
a group assessment problem. In the realm of group assessment problems, researchers have introduced
various assessment methodologies. For instance, scholars (Brauers, & Zavadskas, 2010) proposed
the multi-objective optimization by ratio analysis (MOORA) method, comprising the ratio system
and reference point approaches. Subsequently, they integrated the full multiplication method into
MOORA, forming the MULTIMOORA method. This method yields three distinct results and proves
to be relatively straightforward and robust, making it suitable for addressing group consensus issues.
Due to the excellent stability of MULTIMOORA, it is chosen as the assessment method in this paper.
However, the assessment results derived from these three assessment types may not always align,
necessitating information fusion techniques. Information fusion involves amalgamating diverse data
using specific technical means, enabling effective processing. Among these, the D-S evidence theory
(DSET) (Dempster, 1967; Shafer, 1978) serves as a classic information fusion approach, quantifying
evidence uncertainty via belief and plausibility functions and offering various evidence synthesis
methods to tackle fusion challenges effectively. In recent years, scholars have extensively researched
DSET. Therefore, this paper constructs a stable energy consumption assessment group consensus
decision model through MULTIMOORA and DSET.

3
International Journal on Semantic Web and Information Systems
Volume 20 • Issue 1 • January-December 2024

The Motivations and Contributions of This Paper

Recognizing the significance of industrial energy consumption assessment, particularly in
enhancing sustainability and resource efficiency, the paper seeks to construct a method tailored to
assess industrial energy consumption effectively. This paper aims to explore a robust approach for group
consensus within interval type-2 fuzzy information systems (IT2FISs) using the MULTIMOORA
method. Specifically, operating within the context of IT2FSs, this paper establishes group consensus
and integrates the MULTIMOORA method with DSET, presenting a more resilient decision model and
offering theoretical underpinning. Finally, via instance analysis and comparative examination using
data from the steel industry's energy consumption dataset, this paper showcases our contributions
in four key areas:

• Drawing on the method of maximizing deviations, this paper incorporates DMs' consensus
into the weight calculation process to enhance industrial energy consumption assessment. As
a result, DMs with higher consensus will exert greater influence on the assessment process,
thereby improving the consistency and reliability of decisions regarding energy consumption
in the industry.
• In order to conduct social network analysis more effectively and achieve industrial energy
consumption assessment more efficiently, this paper proposes a novel adjustment strategy. In the
adjustment process, this paper proposes an adjustment strategy based on negative DMs. When
the consensus remains unchanged (set as 10 consecutive adjustments without change in this
paper), we select the DM with the second lowest consensus for adjustment. Subsequently, this
paper continues with the regular adjustment strategy. Additionally, this paper introduces random
parameters to enhance the performance of CRP.
• This paper treats DMs as a social network and develops a group consensus model within the
framework of IT2FSs, employing the MULTIMOORA method and DSET.
• The model's effectiveness and feasibility are validated through practical instances and comparative
analysis of energy consumption in the steel industry.

The framework of this paper is as follows. The next section reviews relevant work and the
following section organizes foundational knowledge. We then proceed with a section that establishes
the group consensus model and a section that applies this model to energy consumption assessment
in the steel industry. Next, we include a section that conducts sensitivity analysis and comparative
analysis and conclude in the final section.

RELEVANT WORK

In recent years, many relevant methods (Feng et al., 2024; Nahar et al., 2023; Piper & Rodger,
2022; Ren et al., 2023) have emerged. As societies worldwide grapple with the challenges of climate
change and sustainable development, understanding and effectively managing energy usage has
become increasingly crucial. Energy consumption not only impacts environmental sustainability
but also influences economic stability and social well-being. In response to these pressing issues,
researchers have developed numerous approaches to assess and mitigate energy consumption across
various sectors and industries, each offering unique insights and opportunities for improving energy
efficiency and reducing environmental impact.

Industrial Energy Consumption

As environmental issues become increasingly severe, more and more scholars are conducting
research in the field of industrial energy consumption. Kalinke et al. (2021) focused on comparing
energy consumption data between industrial facilities and households, assessing the performance gap

4
 International Journal on Semantic Web and Information Systems
Volume 20 • Issue 1 • January-December 2024

of seven load disaggregation algorithms when applied to industrial data versus household data. Hu
et al. (2023) summarized recent research on energy consumption and carbon emissions forecasting
for industrial processes, proposing three novel forecasting models to address challenges in accuracy
and generalizability, offering a potential pathway for establishing more robust forecasting models
in the future. Wu et al. (2023) analyzed energy consumption behavior in a Shanghai industrial park,
employing big data analysis to develop an energy load forecasting model for industrial enterprises,
with an investigation into the impact of meteorological data on forecasting accuracy. Xia et al.
(2022) investigated the periodic driving factors of industrial energy consumption structure (ECS)
in China under stricter environmental regulations, using a multi-layer logarithmic mean Divisia
index decomposition model to analyze ECS trends and identify key drivers for informing energy
and environmental policies. They all conducted research on industrial energy consumption, but they
overlooked the perspective of social network analysis. Therefore, this paper selects industrial energy
consumption as the research background and establishes a group consensus model.

IT2FSs
In recent years, many scholars conducted research in the environment of IT2FSs. Zhou et al.
(2018) developed a novel sorting method for trapezoidal IT2FSs, utilizing the incentre point of fuzzy
sets. Touqeer et al. (2020) utilized the grey relation projection method, a fusion of the “grey relational
analysis method” and the “projection method,” to rank the alternatives. Jin et al. (2023) introduced
three axiomatic definitions for IT2TrF information measures and devised a decision-making method
for supplier selection. Tekeli et al. (2023) conducted a numerical performance analysis of tank
coatings in chemical tankers using a multi-criterion decision-making tool based on IT2FS, aiming
to inform shipowners and safety professionals. Liu et al. (2024) proposed a method for assessing
the failure risk of pipelines in process systems, which combined dynamic Bayesian network (DBN)
with IT2FS, effectively reducing subjectivity and uncertainty in expert opinions and facilitating
preventive measures for pipeline managers. From the literature mentioned above, it becomes evident
that IT2FSs can better capture the uncertainty between individuals. Therefore, this paper conducts
energy consumption assessment based on IT2FSs.

CRP
Group consensus has been a recent research hotspot. Currently, there are two methods for
measuring consensus: (a) consensus measurement based on the distance from collective preferences
and (b) consensus measurement based on the distance between DMs. The main feedback mechanisms
can also be divided into two categories: (a) feedback mechanisms with identification and directional
rules and (b) Feedback mechanisms with minimal adjustments or costs (FMMA/C). Zhang et al.
(2019) proposed a series of comparison criteria to measure CRP efficiency and conducted simulation
experiments. Qin et al. (2022) established the minimum cost consensus model by integrating the
satisfaction index and an optimized CRP, ranking group opinions within each subgroup to determine
the final ranking of alternatives. Zhang et al. (2024) proposed additional adjustments during the
CRP within the context of IT2FSs, establishing a more stable method for consensus reaching. They
applied this method in the agricultural field and obtained favorable assessment results. Rosenberg
et al. (2023) applied CRP in the medical field to assess the treatment status of insomnia, describe
the gaps in current understanding and treatment, determine the current best practices, and provide
updated educational recommendations for primary healthcare providers. From the preceding text, it is
apparent that CRP is more conducive to precise assessment and enables analysis from a social network
perspective. Therefore, this paper utilizes group consensus to fully leverage the diverse perspectives
of multiple DMs and conduct energy consumption assessment more accurately.

5
International Journal on Semantic Web and Information Systems
Volume 20 • Issue 1 • January-December 2024

MULTIMOORA
MULTIMOORA is a relatively robust decision-making method. Hafezalkotob et al. (2019)
provided an overview of MULTIMOORA, summarizing its development and applications. Liu and Li
(2019) integrated the MULTIMOORA method with prospect theory into probabilistic linguistic sets.
Qin and Ma (2022) developed an integrated model based on PT3 and extended MULTIMOORA with
IT2FSs. Akram et al. (2023) successfully developed a proficient MULTIMOORA method utilizing
2-tuple linguistic Pythagorean fuzzy sets. Wang et al. (2024) integrated subjective and objective
weight calculation methods, along with the MULTIMOORA and Borda methods, to comprehensively
assess the ecological governance of the Yellow River Basin and provide policy recommendations.
Zhou et al. (2024) developed an extended MULTIMOORA method to assess and rank offshore wind
power station siting alternatives, exemplified in Shandong Province, China, ensuring reliability and
stability through sensitivity and comparative analyses. In summary, MULTIMOORA is considered a
relatively stable assessment method. Therefore, this paper establishes a more robust decision model
using the MULTIMOORA method to achieve energy consumption assessment.

DSET
DSET has been widely applied in fields such as pattern recognition, artificial intelligence
and decision analysis. Deng (2020) provided a comprehensive review and analysis of uncertainty
measures in evidence theory, focusing on the development and application of Deng entropy, addressing
controversies, and discussing future challenges. Zhang et al. (2021) proposed a decision-making
method based on DSET, extended to a double hierarchy hesitant fuzzy linguistic environment. Xiao
et al. (2022) introduced and explored several generalized evidential divergences to quantify the
disparity and variation among basic belief assignments (BBAs) in DSET. Zhang and Xiao (2023)
presented a sophisticated method for generating BBAs, leveraging the triangular fuzzy number. Wang
et al. (2024) introduced an innovative reliability analysis approach for functionally graded plates
utilizing the ellipsoid-interval-based evidence theory. This method aims to mitigate multi-source
uncertainties in structural safety assessment, particularly emphasizing thermal protection systems.
After the assessment using MULTIMOORA in this paper, it is necessary to utilize evidence theory
for information fusion to obtain the final assessment results. Therefore, this paper utilizes the DSET
for information fusion to obtain the final assessment result.

BASIC KNOWLEDGE

In this section, the basic knowledge relevant to this article will be introduced.

IT2FSs
Due to individual cognitive differences, various DMs may assess uncertain information differently.
IT2FSs, compared to T2FSs, are better equipped to handle such situations. For example, perceptions
of high energy consumption differ among DMs. Some might consider an energy consumption of
1000 kWh to be "high," while others may consider 1000 kWh not to be high. In this case, IT2FSs can
effectively represent the assessments of different DMs regarding whether 1000 kWh is high energy
consumption. Therefore, the membership degree interval for 1000 kWh being considered high energy
consumption may be [0.8, 1].

Definition 1
Setting X as the domain of A, a T2FS A on the domain can be expressed as follows (Mendel et
al., 2006):

|
A = {(x, u), uA(x, u) ∀ x ∈ X, u ∈ Jx ⊆ [0, 1], 0 ≤ uA(x, u) ≤ 1} (1)

6
 International Journal on Semantic Web and Information Systems
Volume 20 • Issue 1 • January-December 2024

where x serves as the primary variable, u as the secondary variable, and the primary membership
function of x is Jx ⊆ [0, 1]. Equation 1 can be represented as follows:

A = ∫ ∫ uA(x, u) / (x, u) (2)

x∈Xu∈Jx

where uA(x, u) = 1 and A are IT2FS.

Definition 2
Considering X as the domain, an IT2FS A can be represented as follows (Mendel et al., 2006):

A = ∫ ∫ 1 / (x, u) (3)
x∈Uu∈Jx

where the meaning of x and u is the same as in Equation 1. It is evident that IT2FSs are a
particular case of T2FSs.

Definition 3
Let A˜ 1 and A ˜ = (A
˜ 2 be two generalized T2F numbers and A ˜ 2) be an IT2F number (Mendel
˜ 1, A
˜
et al., 2006). Then, the IT2F number A defined on the domain X can be represented as follows:

˜ = (A
A ˜ 2) = ((a 1, b 1, c 1, d 1; s 1, t 1), (a 2, b 2, c 2, d 2; s 2, t 2))
˜ 1, A (4)

where the condition a 1 ≤ b 1 ≤ c 1 ≤ d 1, a 2 ≤ b 2 ≤ c 2 ≤ d 2, 0 ≤ s 2 ≤ s 1 ≤ 1, 0 ≤ t 2 ≤

t ≤ 1 is met. In this context, s 1 and t 1 represent the membership degrees of b 1 and c 1 respectively,
1

while s 2 and t 2 represent the membership degrees of b 2 and c 2 respectively. To simplify further
procedures, this paper will clarify the operational regulations of IT2F numbers.
In the context of energy consumption assessment, the detected evaluation values of different
attributes are actual numerical values. To convert these into IT2FSs, we first convert them into
linguistic terms and then into IT2FSs. This conversion relationship is represented through different
scoring functions of IT2FSs. Specifically, detailed explanations of these conversion relationships
are provided in section Assessing Energy Consumption using MULTIMOORA of this paper. The
conversion relationships are as follows:

(0.0,0.103896] -Low (L) - ((0,0.1,0.1,0.3;1,1), (0.05,0.1,0.1,0.2;0.9,0.9));

(0.103896,0.292208] - Medium Low (ML) - ((0.1,0.3,0.3,0.5;1,1), (0.2,0.3,0.3,0.4;0.9, 0.9));

(0.292208,0.5] - Medium (M) - ((0.3,0.5,0.5,0.7;1,1), (0.4,0.5,0.5,0.6;0.9,0.9));

(0.5,0.707792] - Medium High (MH) - ((0.5,0.7,0.7,0.9;1,1), (0.6,0.7,0.7,0.8;0.9,0.9));

(0.707792,0.896104] - High (H) - ((0.7,0.9,0.9,1;1,1), (0.8,0.9,0.9,0.95;0.9,0.9));

7
International Journal on Semantic Web and Information Systems
Volume 20 • Issue 1 • January-December 2024

(0.896104,1] - Very High (VH) - ((0.9,1,1,1;1,1), (0.95,1,1,1;0.9,0.9)).

Example 1. In the assessment of energy consumption in the steel industry, if the normalized
value of the lagging reactive power attribute is 0.6, then, according to the corresponding relationship,
its linguistic term would be “Medium High” (MH). Consequently, the corresponding IT2FS would
be ((0.5,0.7,0.7,0.9;1,1), (0.6,0.7,0.7,0.8;0.9,0.9)).

Definition 4
˜ and A
Assuming that A ˜ are two IT2F numbers, and the following rules are true (Chen & Li,
1 2
2010):

⎜ ⎟
⎛(a11 + a21, b11 + b21, c11 + c21, d11 + d21 ;⎞
min{s11, s21}, min{t11, t21}),
˜ ⊕A
A ˜ = (A ˜ 2) ⊕ ( A
˜ 1, A ˜ 2) =
˜ 1, A (1)
(a1 + a2 , b1 + b2 , c1 + c2 , d1 + d2 ;
1 2 1 1 2 2 2 2 2 2 2 2 2 2

⎝ min{s12, s22}, min{t12, t22}) ⎠

⎜ ⎟
⎛(a11 × a21, b11 × b21, c11 × c21, d11 × d21 ;⎞
min{s11, s21}, min{t11, t21}),
˜ ⊗A
A ˜ = (A ˜ 2) ⊗ ( A
˜ 1, A ˜ 2) =
˜ 1, A (2)
(a1 × a2 , b1 × b2 , c1 × c2 , d1 × d2 ;
1 2 1 1 2 2 2 2 2 2 2 2 2 2

⎝ min{s12, s22}, min{t12, t22}) ⎠

( (k a 2, k b 2, k c 2, k d 2; s 2, t 2) )
˜ = k(A ˜ 2) =
˜ 1, A (k a 1, k b 1, k c 1, k d 1; s 1, t 1),
kA ,k ≥ 0

( (k d 2, k c 2, k b 2, k a 2; s 2, t 2) )
˜ = k(A ˜ 2) =
˜ 1, A (k d 1, k c 1, k b 1, k a 1; s 1, t 1),
kA ,k < 0 (3)

( (( a 2) k, ( b 2) k, (c 2) k, (d 2) k; s 2, t 2) )
˜ k = (A ˜ 2) k =
˜ 1, A
( (a 1) k, (b 1) k, (c 1) k, ( d 1) k; s 1, t 1) ,
A ,k ≥ 0 (4)

⎜ ⎟
⎛(a11 − d21, b11 − c21, c11 − b21, d11 − a21 ;⎞
min{s11, s21}, min{t11, t21}),
˜ ⊖A
A ˜ = (A ˜ 2) ⊖ ( A
˜ 1, A ˜ 2) =
˜ 1, A (5)
(a1 − d2 , b1 − c2 , c1 − b2 , d1 − a2 ;
1 2 1 1 2 2 2 2 2 2 2 2 2 2

⎝ min{s12, s22}, min{t12, t22}) ⎠

8
 International Journal on Semantic Web and Information Systems
Volume 20 • Issue 1 • January-December 2024

⎜ ⎟
⎛a11 / d21, b11 / c21, c11 / b21, d11 / a21 ;⎞
min(s11, s21), min(t11, t21)
˜ ⊘A
A ˜ = (A ˜ 2) ⊘ ( A
˜ 1, A ˜ 2) =
˜ 1, A (6)
1 2 1 1 2 2
a12 / d22, b12 / c22, c12 / b22, d12 / a22 ;
⎝ min(s12, s22), min(t12, t22) ⎠

In group consensus, comparing two fuzzy numbers is essential. Hence, the score function
pertaining to IT2F numbers is established below.

Definition 5
The score function of A ˜ = (A
˜ can be defined below, where A ˜ 2) = ((a 1, b 1, c 1, d 1; s 1, t 1), (a 2,
˜ 1, A
b , c , d ; s , t )) is an IT2F number (Hu et al., 2013).
2 2 2 2 2

˜) = _1 (a 1 + a 2 + b 1 + b 2 + c 1 + c 2 + d 1 + d 2) × _1 (s 1 + t 1 + s 2 + t 2)
s (A (5)
8 4

In 2013, Hu et al. (2013) introduced the concept of distances between two IT2F numbers. To
compute the weight calculation, we provide the distance formula below.

Definition 6
˜ and A
If A ˜ represent two IT2F numbers, the formula for calculating their distance is below
1 2
(Hu et al., 2013):
_
D (A
˜ ,A
1
˜ ) = √D + D
2 1 2

D1 = (a12 − a22) 2 + (b12 ⋅ s12 − b22 ⋅ s22) 2 + (c12 ⋅ t12 − c22 ⋅ t22) 2 + (d12 − d22) 2

D2 = (a11 − a21) 2 + (b11 ⋅ s11 − b21 ⋅ s21) 2 + (c11 ⋅ t11 − c21 ⋅ t21) 2 + (d11 − d21) 2 (6)

CRP
In order to satisfy the majority of DMs, CRP has been incorporated into assessment frameworks
by numerous scholars. Existing CRP models typically adhere to a common framework comprising
four primary procedures (Zhang et al., 2020):

1. DMs articulate their original or modified preferences through a designated format for preference
presentation.
2. Next, an aggregation function is utilized to merge individual preferences into a unified collective
preference.

9
International Journal on Semantic Web and Information Systems
Volume 20 • Issue 1 • January-December 2024

Figure 2. The chart of the CRP framework

3. DMs actively employ a consensus measurement approach to gauge the degree of consensus
among themselves.
4. The feedback mechanism actively generates suggestions for modifying preferences.

Iterations of 1–4 continue until the predefined consensus degree is achieved. For ease of
understanding, Figure 2 illustrates the CRP framework.
Generally, consensus emerges as DMs adjust their preferences towards mutually agreed-upon
points. Mechanisms for adjustments play a crucial role in recommending preference modifications.
Two categories of adjustment mechanisms typically guide preference modifications within the current
body of literature:

• Feedback mechanisms incorporating identification and direction rules. Identification rules

pinpoint DMs, alternative solutions, and preference values exhibiting poor consistency, thus
warranting involvement in the preference modification process. To facilitate the achievement of
group consensus, direction rules guide preference modifications.
• Adjustment mechanism with FMMA/C, designed to reduce disparities between original
preferences and adjusted ones, or to minimize the overall cost of achieving consensus.

THE CRP IN SOCIAL NETWORK ANALYSIS

The CRP in a social network of DMs aims to more accurately monitor the level of steel energy
consumption over different days. We established a group consensus model for this purpose. However,
before making decisions, we need to achieve group consensus among three decision makers (selecting
three different time points each day).
This section will explore approaches to CRP within an IT2F framework. The procedure consists
of two phases: (a) computing the consensus degree measurement method and (b) executing a feedback
adjustment process.

10
 International Journal on Semantic Web and Information Systems
Volume 20 • Issue 1 • January-December 2024

The Consensus Degree Calculation

Within this section, we investigate approaches to assessing consensus. Concerning the correlation
between consensus measurement and consensus adjustment discussed in the subsequent section, the
former serves as the foundation for the latter. Incorporating both elements facilitates the group in
reaching its final decision.
This paper calculates the consensus degree between every pair of DMs, gradually aggregating
to ascertain the consensus degree of the group. This method guarantees both a substantial consensus
among individual DMs and a considerable consensus degree across the group. By employing this
method, we establish a consensus model that effectively mitigates instances where individual experts
deviate from the group consensus, even if the overall consensus level is high.
The procedure for computing the group consensus value involves the following steps:

1. The consensus level between the DM pc and the DM pv(c, v = 1, 2, ⋯ ,l) is calculated as indicated
below. It is noteworthy that the distances have been normalized in this paper.

n m

mn ∑∑d(Rc(xj, yk), Rv(xj, yk))

1
EL(pc, pv) = 1 − _ (7)
j=1 k=1

2. The consensus level between the DM pc and all other DMs is calculated as follows:

l − 1 i=1,i≠c( mn d(Ri(xj, yk), Rc(xj, yk)))

1
AL(pc) = _ 1
∑ 1−_ (8)

3. Ultimately, the group consensus level GL can be computed as follows:

l(l − 1) i=1 i=1,i≠c( mn ∑∑d(Ri(xj, yk), Rc(xj, yk)))

l l n m
1
GL = _ 1
∑∑ 1−_ j=1 k=1
(9)

Using Equations (7)–(9), by assessing the group consensus degree, we can ascertain the level
of agreement among DMs. It is worth noting that the consensus level achieved by DMs may vary
depending on the specific assessment scenarios.

Adjusting the Value of Consensus Degrees

In facing an assessment issue, establishing a threshold for consensus degree, denoted as GL ‾, is
crucial to determine an acceptable level of consensus. When the computed whole consensus value
falls beneath the predetermined threshold, adjustment procedures are initiated. DMs are required to
modify their assessment values according to the adjustment rules until the whole consensus value
meets the predefined threshold. Next, we will explore the intricate details of the adjustment rules.
Upon computing the consensus degree GL, if GL ≥ GL ‾, the DMs remain unchanged. However,
if GL < GL ‾, this paper pinpoints the DM whose impact on the group consensus degree is minimal,
signaling a significant deviation in their assessment value compared to others. The formula is outlined
below.

pt = min c(AL(pc)) (10)

11
International Journal on Semantic Web and Information Systems
Volume 20 • Issue 1 • January-December 2024

If all the assessment values of a DM are adjusted once the DM pt is identified, the original
assessment data may not be effectively retained. Consequently, we need to assess which assessment
data provided by the DM pt should be adjusted.
Where Rt(xj, yk) represents the assessment measure of the t-th DM concerning the k-th attribute
in the j-th alternative, and RC(xj, yk) represents the collective assessment value matrix concerning the
k-th attribute in the j-th alternative.
The greater the difference between the appraisal by the DM regarding the attribute yk in the
alternative xj and its counterpart in the collective matrix, the more distant it is from the consensus;
conversely, the smaller the gap, the closer to the consensus it is. Hence, we prioritize selecting the
assessment by the DM pt concerning the attribute yk in the alternative xj, which demonstrates the
greatest deviation from its associated collective decision matrix, for adjustment.
When we select the DM pt to be modified and identify the assessment value for modification,
we can proceed with the adjustment. The precise adjustment formula is below.

R ' t(xj, yk) = r1jk Rt(xj, yk) ⊕ r2jk RC(xj, yk) (11)

r1jk and r2jk are the two parameters introduced to adjust the assessment value, computed as follows:

Rt(xj, yk)
r1jk = 1 − _ n * ra

r2jk = ra − r1jk (12)

In order to make our adjustment method applicable across different scenarios, we integrate
a random adjustment parameter, denoted as ra within the range of 0 to 1. The strategy assists in
identifying the most effective adjustment parameters, thereby bolstering the consensus level among
DMs.
To tackle the problem of limited accuracy in program results hindering consensus enhancement,
we implement an adjustment strategy based on negative DMs. This strategy involves dividing the
adjustment process into two stages. During the initial stage of assessment matrix adjustment, should
the consensus level persist unchanged following 10 consecutive adjustments, the time point ph is then
subject to adjustment. Here, ph represents the time point characterized by the second lowest level of
consensus. The adjustment approach employed in this stage closely resembles the initial one. This
particular tactic is termed as a negative DMs-based adjustment strategy, with the random parameter
designated as r a2.

Assessing Energy Consumption using MULTIMOORA

Once a consensus has been reached among the DMs, we move on to the subsequent processing
of their values. We standardize the data by transforming each attribute value into its respective
linguistic scale. Subsequently, we translate the linguistic scale into the compatible IT2F number. The
corresponding relationships derived in this article are as follows:
When the normalized value is 0, the corresponding linguistic scale is “Very Low” (VL), and the
corresponding IT2FS is ((0,0,0,0.1;1,1), (0,0,0,0.05;0.9,0.9)). This correspondence is represented in the
article as '0 - Very Low (VL) - ((0,0,0,0.1;1,1), (0,0,0,0.05;0.9,0.9))'. That means, we can also obtain:

12
 International Journal on Semantic Web and Information Systems
Volume 20 • Issue 1 • January-December 2024

(0.0,0.103896] - Low (L) - ((0,0.1,0.1,0.3;1,1), (0.05,0.1,0.1,0.2;0.9,0.9));

(0.103896,0.292208] - Medium Low (ML) - ((0.1,0.3,0.3,0.5;1,1), (0.2,0.3,0.3,0.4;0.9, 0.9));

(0.292208,0.5] - Medium (M) - ((0.3,0.5,0.5,0.7;1,1), (0.4,0.5,0.5,0.6;0.9,0.9));

(0.5,0.707792] - Medium High (MH) - ((0.5,0.7,0.7,0.9;1,1), (0.6,0.7,0.7,0.8;0.9,0.9));

(0.707792,0.896104] - High (H) - ((0.7,0.9,0.9,1;1,1), (0.8,0.9,0.9,0.95;0.9,0.9));

(0.896104,1] - Very High (VH) - ((0.9,1,1,1;1,1), (0.95,1,1,1;0.9,0.9)).

Subsequently, we acquire DMs' weights wi and the attributes' weights μik. Considering the
diversity in consensus levels among different DMs, we integrate the previous method of computing
DMs' consensus degree AL(pi) with the deviation maximization technique to ascertain the weight
wi of DMs. If DM pi assess the attribute yk concerning the alternative xj as Ri(xj, yk), then the DM's
weight wi can be represented as shown below. It is noteworthy that initially, the DMs' weight is wi'.

w ' i + AL ' i
wi = ____________
l l
∑w ' i + ∑AL ' i
i=1 i=1

_______________
l m n
1
∑ ∑∑d(Ri(xj, yk), Rh(xj, yk))
w ' i = _____________________
l
h=1,h≠i k=1 j=1

1
∑______________________
l m n
i=1 ∑ ∑∑d(Ri(xj, yk), Rh(xj, yk))
h=1,h≠i k=1 j=1

AL(ei)
AL ' i = _
l (13)
∑AL(ei)
i=1

At this juncture, the attribute weight μik in each decision matrix can be computed by maximizing
the deviation method and is expressed as follows.

13
International Journal on Semantic Web and Information Systems
Volume 20 • Issue 1 • January-December 2024

Table 1. The BBAs of different assessment results

BBAs x1 ⋯ xj ⋯ xn

M1 as 1 1 ⋯ as 1 j ⋯ as 1 n

M2 as 2 1 ⋯ as 2 j ⋯ as 2 n

M3 as 3 1 ⋯ as 3 j ⋯ as 3 n

m m
∑∑(d(Ri(xj, yk), Ri(xl, yk)))
uik = ______________________
j=1 l=1
n m m (14)
∑∑∑ l=1
(d(Ri(xj, yk), Ri(xl, yk)))
k=1 j=1

The alternatives were subsequently assessed employing three MULTIMOORA techniques,

namely, as 1 j, as 2 j, and as 3 j. The greater the values of as 1 j and as 3 j, the more preferable the option
j.The smaller the value of as 2 j, the more advantageous the plan j. Subsequently, the assessment
values are organized into rankings, leading to the creation of three sets of rankings. Below are the
three methodologies for assessment:

as 1 j = ∑wi(∑μk R(xj, yk) − ∑ μk R(xj, yk))

l g m
(15)
i=1 k=1 k=g+1

k { ( k ( j k)
as 2 j = ∑wi max d μ R x , y , μk R'(xj, yk))} (16)
i=1

l ∏(R(xj, yk)) μ k

as 3 j = ∑wi ___________
k=1
m (17)
i=1 ∏ (R(xj, yk)) μ k

k=g+1

Next, the aforementioned three sorting values are arranged in ascending order to obtain three
types of sorting results.

The Fusion of Assessment Results

After assessing the options, it is imperative to amalgamate the three assessment results derived via
MULTIMOORA to obtain the ultimate assessment. Should the calculated assessment for an alternative
align across all three methods, the final assessment remains unchanged. However, if discrepancies
arise among the assessments, conflict resolution becomes necessary for the ultimate decision. In this
study, DSET is employed to address conflicting assessments effectively.
In 2021, a belief function evidence correlation coefficient (ECC) was proposed by Xiao and Cao
(2021) to assess the conflict between two pieces of evidence. Initially, we compile a table featuring
the alternatives xj alongside their respective tripartite assessment outcomes, referred to as as 1 j, as
2 j, and as 3 j. Additionally, we denote the assessment values obtained from the three methods as M1,
M2, and M3. This is illustrated in Table 1.
Subsequently, we address the conflict through the following approaches.

14
 International Journal on Semantic Web and Information Systems
Volume 20 • Issue 1 • January-December 2024

1. Construct the correlation matrix through ECC.

⎢ ⎥
⎡ECC(M1, M1) ECC(M1, M2) ECC(M1, M3)⎤
matri xECC = ECC(M2, M1) ECC(M2, M2) ECC(M2, M3) (18)
⎣ECC(M3, M1) ECC(M3, M2) ECC(M3, M3)⎦

2. Determine the support of asq(q = 1, 2, 3) using the following method.

3
SD(asq) = ∑ ECC(asi, asq) (19)
i=1|i≠q

3. Calculate the trustworthiness of A using the following method.

SD(asq)
CD(asq) = _
3 (20)
∑SD(asq)
q=1

4. Compute the weighted average evidence (WAE) in the manner described below.

3
WAE(xj) = ∑CD(asq) × as qj (21)
q=1

Algorithm 1 is the algorithmic process of the decision model proposed in this paper.

Algorithm 1
The Algorithmic Process of the Decision Model
Input: The decision matrix
Output: The optimal alternative x*
1. for i=1 to l do
2. for j=1 to n do
3. for k=1 to m do
4. Normalize the assessment values.
5. Convert the assessment values into IT2FSs.
6. end // Data preprocessing.
7. for i=1 to l do
8. for j=1 to n do
9. for k=1 to m do
10. Calculate DMs’ weights via Eq. (13).
11. Calculate attribute weights via Eq. (14).
12.end // Calculate weights.
13.for i=1 to l do
14. for j=1 to n do
15. for k=1 to m do
16. Adjust the random parameters value within the range of 0 -
1 to find the highest consensus degree.

15
International Journal on Semantic Web and Information Systems
Volume 20 • Issue 1 • January-December 2024

17. Select the optimal values for raand ra2, as well as the
most suitable consensus threshold GL¯.
18.end //Determine the values of the parameters and
the consensus threshold GL¯.
19. for i=1 to l do
20. for j=1 to n do
21. for k=1 to m do
22. Initiate the CRP procedure via Eqs. (7) - (9).
23.end // Calculate the group consensus degree GL.
24.if GL≥GL¯ do
25. DMs reach a consensus and make an assessment.
26.else do
27. Adjust using Eqs. (10) - (12).
28.end //Facilitate DMs to reach a consensus.
29. for i=1 to l do
30. for j=1 to n do
31. for k=1 to m do
32. Utilize the MULTIMOORA method for decision-making via Eqs.
(15) - (17).
33.end // Make assessment using MULTIMOORA.
34. for i=1 to l do
35. for j=1 to n do
36. for k=1 to m do
37. Fuse the three decision results using evidence theory via
Eqs. (18) - (21).
38.end // Obtain the final assessment results.
39. Get the optimal alternative x*.
For easy understanding, this paper makes a complete flowchart of the algorithm as shown in
Figure 3.

A CASE STUDY

In the following, we apply the IT2FS-based group consensus model proposed in this paper to
the energy consumption dataset in the steel industry. This paper will detail the precise process of
consensus-reaching and evaluation, culminating in the presentation of the final assessment outcomes.
The dataset is sourced from a steel company in South Korea, containing energy consumption
data detected every 15 minutes each day. In this paper, data from 7:00, 12:00, and 19:00 on 15 days
in 2018 are selected for experimental analysis.

Data Preprocessing
Because there is currently no dataset featuring IT2F numbers, we must first convert our data to this
format before utilizing the established model for practical applications. Prior to this transformation,
it is necessary to preprocess the dataset.
In the context of an IT2FIS (U, V, Ri, S), this section focuses on initially normalizing the values
of each attribute within the dataset to facilitate subsequent calculations. Let V = {y1, ⋯ , yk, ⋯ , ym
} denote the attribute set.
yk − ymin
yk' = _
y −y (22)
max min

16
 International Journal on Semantic Web and Information Systems
Volume 20 • Issue 1 • January-December 2024

Figure 3. The flowchart of the industrial energy consumption assessment algorithm

Following the conversion rules mentioned in Section 4.3, we converted real numbers into IT2FSs.
Example 2. Assume that during the energy consumption assessment, the attribute yk has three
values: z1 = 30, z2 = 42, and z3= 50. After normalization, z2 has a value of 0.6. According to the
conversion rules mentioned in section Assessing Energy Consumption using MULTIMOORA, this
value falls within the linguistic term “MH” range, corresponding to the IT2FS ((0.5, 0.7, 0.7, 0.9; 1,
1), (0.6, 0.7, 0.7, 0.8; 0.9, 0.9)).

The Application of Group Consensus Models in Energy

Consumption Within the Steel Industry
Assessing industrial energy consumption, especially in the steel industry, is vital due to its
significant contribution to overall energy usage and its impact on production costs, competitiveness,
and environmental sustainability. Optimizing energy efficiency in steel production not only reduces
operational expenses but also ensures resilience against market fluctuations and supports long-term
viability in a transitioning energy landscape.

17
International Journal on Semantic Web and Information Systems
Volume 20 • Issue 1 • January-December 2024

Figure 4. The chart of partial parameter adjustment

The dataset (Sathishkumar et al., 2023) used in this section is sourced from a steel company in
South Korea, comprising energy consumption data detected every 15 minutes each day. In this paper,
data from 7:00, 12:00, and 19:00 on 15 days are selected for experimental analysis.
U = {x1, x2, ⋯ , x15} represents the set of candidate dates, and V = {y1, y2, ⋯ , y6} represents the
attribute set, where y1 denotes industrial energy consumption, y2 represents lagging reactive power, y3
represents leading reactive power, y4 denotes carbon dioxide emissions, y5 represents lagging power
factor, and y6 denotes leading power factor.

The Process of Consensus Among DMs

Based on the aforementioned conditions, this paper applies the group consensus decision-making
model constructed to the energy consumption in the steel industry. Firstly, we need to input the IT2FSs
decision matrix. Before making the decision, we need to set a consensus threshold. Since directly
setting the consensus threshold is subjective, we perform 500 consensus adjustments on the input
data and choose the higher consensus value as the consensus threshold for this decision.
Next, we experimentally test the values of random parameters ra and r a2, which range from [0,1]
with a step size of 0.01. After testing, it was found that when ra is 0.99 and r a2 is 0.9, the highest
consensus value obtained is 0.5475354354011809. Figure 4 shows the variation diagram of different
values of ra when parameter r a2 is set to 0.9.
‾ and initiate the CRP. According to Equations
We then select 0.54 as the consensus threshold GL
(14)–(19), group consensus degree calculation and adjustment process are carried out. Tables 2 and
3 demonstrate the initial group consensus degree calculation process.
Tables 2 and 3 display the intermediate process of calculating the group consensus degree
value GL. Specifically, Table 2 shows the consensus degree values between each pair of DMs (in
this instance, the DMs are time points). Table 3 presents the consensus degree value for each DM.

18
 International Journal on Semantic Web and Information Systems
Volume 20 • Issue 1 • January-December 2024

Table 2. The consensus degree between each pair of DMs

Description Calculation Results

The consensus value between 7:00 and 12:00 0.4903248313974802
The consensus value between 7:00 and 19:00 0.5239236503231285
The consensus value between 12:00 and 19:00 0.5144558016122693

Table 3. The consensus degree results for each DM

Description Calculation Results

The consensus value of 7:00 0.5071242408603043
The consensus value of 12:00 0.5023903165048748
The consensus value of 19:00 0.519189725967699

Figure 5. The chart of MULTIMOORA assessment results

The Energy Consumption Assessment Process

After consensus is reached among different DMs, the energy consumption assessment begins.
Initially, the assessment results are calculated using the MULTIMOORA method, as depicted in
Figure 5.
It is worth noting that both as 1 j and as 3 j indicate that higher assessment values correspond to
higher energy consumption, while as 2 j indicates that lower assessment values correspond to higher

19
International Journal on Semantic Web and Information Systems
Volume 20 • Issue 1 • January-December 2024

Table 4. The SD and CD values for different methods

asq SD(asq) CD(asq)

as1 3.9697116678067514 0.36173416747369647

as2 3.489308134822015 0.3179580984294508
as3 3.5150933023916746 0.32030773409685276

energy consumption. Figure 5 illustrates that in the assessment results obtained from as 1 j and as 3 j,
the assessment value on the sixth day is the smallest, indicating the lowest energy consumption. In
as 2 j, the assessment value on the sixth day is the largest, indicating the highest energy consumption.
Next, the energy consumption assessment results are fused using evidence theory. In this process,
the constructed ECC matrix is as follows.

2.4720801994758403 2.0184450306511543 1.9512666371555973

[1.7585715173080985 1.7565217850835764 1.7724419234235478]
matri xECC = 1.7458797454792592 1.7458797454792592 1.743428389342756

The calculation results for SD(asq) and CD(asq ) (q = 1, 2, 3) are shown in Table 4.
Finally, we can obtain the WAE for each alternative date, and the final results are as shown in
Figure 6.
Figure 6 shows that the assessment value obtained on the sixth day is the lowest, indicating the
minimum energy consumption. Therefore, the sixth day is the optimal alternative in this experiment.

EXPERIMENTAL ANALYSIS

In the previous section, this paper validated the established model through examples of steel
energy consumption and obtained the final assessment results. In this section, sensitivity analysis
and comparative analysis will be conducted on the established model.

Sensitivity Analysis
We will experimentally test results by modifying the parameters employed in the model to
compare evaluation outcomes obtained across various parameter settings. This paper will adjust the
consensus threshold GL ‾, adjustment strategies triggering the adjustment strategy based on negative
in the adjustment strategy, as well as the random parameters ra and r a2.
In this paper, adjustments were made to the consensus threshold GL ‾, setting three values 0.543,
0.537, and 0.535, for comparison with the originally proposed consensus threshold. The comparative
results are shown in Figure 7(a). From the graph, it can be observed that when GL ‾ is set to 0.543,
there is a slightly larger fluctuation in the assessment value on the 10-th day. However, in other cases,
the fluctuation in assessment values is minimal and will not affect the assessment results.
Additionally, this paper adjusts the trigger frequency in the adjustment strategy based on negative
feedback. The trigger frequency is modified to 8 times, 6 times, and 3 times, and their assessment
results are compared, as shown in Figure 7(b). The graph indicates that adjusting the trigger frequency
results in slight changes in the evaluation values, but it does not affect the final assessment outcome.

20
 International Journal on Semantic Web and Information Systems
Volume 20 • Issue 1 • January-December 2024

Figure 6. The chart of final assessment results

Finally, this paper also tests the random parameters ra and r a2. In addition to the parameter
values set in this paper, three additional values are set and evaluated. The results are shown in Figure
7(c). It is evident that adjusting the random parameters also does not affect the evaluation results.
In conclusion, the assessment model established in this paper demonstrates a certain level of
stability.

Simulation Analysis
In this section, we delve into a simulation analysis of the model proposed in this paper. The
experiments are conducted using PyCharm 2023.1 on a computer equipped with an Intel Core
i7-10700U 2.90 GHz CPU and 16 GB of RAM.
Initially, we simulate the potential impacts that external factors could impose on the experimental
data, such as abrupt increases or decreases in attribute values due to policy influences. This study
postulates the ramifications of sudden shifts in the second attribute, y2 (lagging reactive power),
prompted by policy changes. We simulate three scenarios: where the value of y2 is halved, doubled,
and quadrupled from its original value. Through this simulation experiment, we observe that across
these scenarios, the assessment results and decision outcomes consistently align with the original
findings. The simulation experiment results are depicted in Figure 8.
The legend in Figure 8 represents the multiples by which the second attribute value has been
modified.
This indicates that our model exhibits robustness to such changes. Even when there are significant
shifts in energy consumption patterns due to external factors, the model's evaluation results remain

21
International Journal on Semantic Web and Information Systems
Volume 20 • Issue 1 • January-December 2024

Figure 7. The chart of sensitivity test results

stable. This robustness is crucial for ensuring reliable decision support in real-world applications,
allowing the model to maintain its performance across various scenarios.
Then, to investigate the impact of large-scale datasets on the assessment, we conducted two
simulation experiments, examining the influence of the number of DMs on the assessment:

• We studied the effect of the number of DMs on the runtime of the experiments.
• We explored the impact of the number of DMs on the rounds of the CRP.

Integrating the analysis of large-scale datasets into these experiments, we obtained two result
charts, as illustrated in Figures 9(a) and (b).

Comparative Analysis
In this section, our model will be compared and analyzed against other assessment models. Three
classic assessment models will be selected for comparison. MULTIMOORA and consensus will be
integrated with these classic assessment models respectively, and the final assessment results will
be compared.
The comparison of assessment results between our model and three classic models is shown
in Figure 10. The graph reveals that, except for the TODIM model, all models indicate the optimal
alternative date as the sixth day. On this day, the assessment values are the lowest, indicating the

22
 International Journal on Semantic Web and Information Systems
Volume 20 • Issue 1 • January-December 2024

Figure 8. The chart of simulation results of external factors impact on model outcomes

least energy consumption. However, in the assessment results of the TODIM model, the sixth day is
the second-best alternative date, with a very small difference in assessment values compared to the
optimal alternative date.
The discrepancy in TODIM results may be attributed to the unique decision-making approach
of TODIM, which emphasizes diversification measures and reference point utilization. However,
this approach may introduce uncertainty in the assessment results, as it ranks alternatives based on
reference points rather than directly on energy consumption performance.
The results of comparing the model combining MULTIMOORA with three classic assessment
models with our model are shown in Figure 11. It can be seen that the assessment result of the
“MULTIMOORA+TODIM Model” indicates the optimal alternative date as the first day, with the
second-best alternative date being the sixth day, and the assessment results on these two days are
very close. Additionally, the optimal alternative date for our model and other methods is the sixth
day, with the second-best alternative date being the first day.
Finally, after combining the group consensus model with three classic assessment models and
comparing them with our model, as shown in Figure 12, it can be observed that the sixth day is the
optimal alternative date in our model, the “Consensus +TOPSIS Model,” and the “Consensus+CFWG
Model .” This indicates that the energy consumption is minimal on the sixth day out of the 15 days
considered. However, in the “Consensus +TODIM Model,” the sixth day is the second-best alternative
date, but the difference in assessment values between the optimal alternative date is minimal,
approximately around 0.1.
The comparative analysis in Figure 12 indicates that our model is feasible and stable.

23
International Journal on Semantic Web and Information Systems
Volume 20 • Issue 1 • January-December 2024

Figure 9. The chart of simulation results of external factors impact on model outcomes

24
 International Journal on Semantic Web and Information Systems
Volume 20 • Issue 1 • January-December 2024

Figure 10. The chart of comparison between our model and classic models

Management Discussion
This paper proposes a comprehensive approach for accurately assessing industrial energy
consumption, incorporating IT2FSs, social network analysis, and information fusion techniques.
This method aims to address uncertainties and subjective biases in industrial energy consumption
assessment, providing essential theoretical and methodological support for energy reduction and
efficiency improvement.
In this study, IT2FSs are employed as a fuzzy environment to better address uncertainties in
assessing industrial energy consumption. Concurrently, through social network analysis, DMs are
conceptualized as a social network to facilitate consensus-reaching among them. This approach aids
in mitigating individual DMs' subjective biases, thereby enhancing the objectivity and fairness of the
assessment. Moreover, the study opted for the more stable MULTIMOORA method for assessment
and utilized evidence theory for information fusion to obtain the final assessment result.
In terms of application, the comprehensive assessment method proposed in this paper offers
significant benefits to the industrial sector, particularly in industries that are major energy consumers,
such as steel production. The steel industry is one of the sectors with substantial energy consumption,
and accurate assessment and optimization of energy consumption are crucial for improving
production efficiency and reducing costs. The method proposed in this paper can provide the steel
industry with more accurate and objective energy consumption assessment results, thereby offering
theoretical and practical support for steel companies to develop scientifically sound energy-saving
and emission-reduction strategies. Additionally, by enabling steel companies to optimize their energy
consumption, the proposed method contributes to reducing production costs (Zhang et al., 2023)

25
International Journal on Semantic Web and Information Systems
Volume 20 • Issue 1 • January-December 2024

Figure 11. The chart of comparison of the assessment results between our model and the “MULTIMOORA+Classic Model”

and enhancing overall economic efficiency (Pettinger et al., 2023) in the steel industry. Our model's
adaptability allows it to handle uncertainty and fuzziness across various high-energy-consuming
industries by adjusting model parameters to fit specific evaluation criteria. We also intend to explore
the application of this assessment model in additional fields (Sangaiah et al., 2023; Xu et al., 2023)
such as healthcare (He et al., 2022; Tan et al., 2024) and transportation (Cao et al., 2022; Chen et al.,
2024), where efficient resource allocation and sustainability efforts are equally vital.

CONCLUSION

This paper utilizes IT2FSs to analyze assessment from a social network perspective, facilitating
group consensus among DMs. It then integrates the MULTIMOORA method with evidence theory
to establish a robust group consensus model based on IT2F evidence fusion. Using the energy
consumption dataset from the UCI database in the steel industry, the effectiveness and feasibility of
the proposed assessment model are validated by assessing different alternative dates. Finally, through
sensitivity and comparative analyses, the stability of the proposed assessment model is confirmed by
comparing assessment results obtained by different models on the same dataset.
Future research can explore both theoretical and practical developments. In theoretical terms,
we can conduct research from the following perspectives:

• Further research directions may explore the refinement and enhancement of IT2FSs to better
capture uncertainties and complexities in decision-making environments. This could involve

26
 International Journal on Semantic Web and Information Systems
Volume 20 • Issue 1 • January-December 2024

Figure 12. The chart of comparison of our model and the “Consensus+Classic Model”

developing more advanced fuzzy logic techniques, employing hybrid fuzzy systems, or leveraging
higher-order fuzzy sets to increase the descriptive power of IT2FSs.
• Integrating IT2FSs with other decision-making theories, such as Bayesian networks, DSET,
or rough set theory, can improve the robustness and reliability of assessment models. This
multidisciplinary approach can help create more comprehensive models that handle a wider
range of uncertainties and provide more nuanced insights.
• The development of new algorithms or the optimization of existing ones to efficiently process and
analyze fuzzy data is another potential research direction. Enhancing computational efficiency
and scalability will be crucial for applying IT2FSs to large and complex datasets.
• Combining IT2FSs with machine learning and AI can lead to adaptive and intelligent systems
capable of learning from data and improving over time, thereby developing predictive models
that provide more accurate and timely decision support.

In practical terms, we can conduct research from the following perspectives:

• There is potential to extend the proposed assessment model to other industrial sectors beyond
steel production. For instance, integrating coal consumption data could offer insights into energy
consumption patterns from an industrial data-driven perspective. This extension could benefit
industries such as chemical engineering, metallurgy, automotive manufacturing, and more. By
incorporating various types of energy consumption data, such as electricity, natural gas, and

27
International Journal on Semantic Web and Information Systems
Volume 20 • Issue 1 • January-December 2024

renewable sources, the model can provide a holistic view of energy usage, helping to identify
inefficiencies and optimize energy consumption across different industrial processes.
• Exploring the application of the model in conjunction with emerging technologies such as Internet
of things (IoT), sensors or machine learning algorithms could enhance its capabilities in real-time
energy consumption monitoring and decision support. IoT sensors can provide continuous data
streams, while machine learning algorithms can analyze this data in real time to detect anomalies,
predict future trends, and recommend optimization strategies.
• Expanding the assessment model into new domains also has wide-ranging applications. Applying
the assessment model within smart manufacturing systems can significantly improve energy
efficiency and production optimization. By integrating with advanced manufacturing technologies,
the model can help achieve a more sustainable and cost-effective production process.
• The model can be applied to green energy management, aiding various industries in transitioning
to renewable energy sources and reducing their carbon footprint. By providing detailed insights
into energy consumption patterns, the model can support the implementation of sustainable
energy practices. Using the model for predictive maintenance can help anticipate equipment
failures and schedule timely maintenance activities, reducing downtime and enhancing energy
efficiency. Real-time monitoring data combined with predictive analytics can provide actionable
insights to maintain optimal operational conditions.

By focusing on these development points, future research could contribute to the broader
adoption and implementation of advanced assessment models in various industrial domains, fostering
more efficient and sustainable energy management practices. This comprehensive approach to both
theoretical and practical development will pave the way for innovations that address the complexities
and challenges of modern industrial energy management.

COMPETING INTERESTS

The authors of this publication declare there are no competing interests.

FUNDING

This paper was supported by the Special Fund for Science and Technology Innovation Teams
of Shanxi (202204051001015), the Training Program for Young Scientific Researchers of Higher
Education Institutions in Shanxi, the Cultivate Scientific Research Excellence Programs of Higher
Education Institutions in Shanxi (CSREP) (2019SK036), and the Wenying Young Scholars of Shanxi
University.

PROCESS DATES

July 23, 2024

Received: May 5, 2024, Revision: June 16, 2024, Accepted: July 243, 2024

CORRESPONDING AUTHOR

Correspondence should be addressed to Chao Zhang (China, [email protected])

28
 International Journal on Semantic Web and Information Systems
Volume 20 • Issue 1 • January-December 2024

REFERENCES

Akram, M., Khan, A., & Ahmad, U. (2023). Extended MULTIMOORA method based on 2-tuple linguistic
Pythagorean fuzzy sets for multi-attribute group decision-making. Granular Computing, 8(2), 311–332. 10.1007/
s41066-022-00330-5
Alrumaih, T. N., & Alenazi, M. J. (2024). CGAAD: Centrality-and graph-aware deep learning model for detecting
cyberattacks targeting industrial control systems in critical infrastructure. IEEE Internet of Things Journal,
11(13), 24162–24182. 10.1109/JIOT.2024.3390691
Bai, W., Ding, J., & Zhang, C. (2020). Dual hesitant fuzzy graphs with applications to multi-attribute decision
making. International Journal of Cognitive Computing in Engineering, 1, 18–26. 10.1016/j.ijcce.2020.09.002
Bai, W., Zhang, C., Zhai, Y., & Sangaiah, A. K. (2024). A robust incomplete large-scale group decision-making
model for metaverse metro operations and maintenance. Applied Soft Computing, 156, 111472. Advance online
publication. 10.1016/j.asoc.2024.111472
Brauers, W. K. M., & Zavadskas, E. K. (2010). Project management by MULTIMOORA as an instrument for
transition economies. Technological and Economic Development of Economy, 16(1), 5–24. 10.3846/tede.2010.01
Cao, H., Garg, S., Kaddoum, G., Hassan, M. M., & AlQahtani, S. A. (2022). Intelligent virtual resource allocation
of QoS-guaranteed slices in B5G-enabled VANETs for intelligent transportation systems. IEEE Transactions on
Intelligent Transportation Systems, 23(10), 19704–19713. 10.1109/TITS.2022.3178267
Chen, H., Lin, M., Liu, J., Yang, H., Zhang, C., & Xu, Z. (2024). NT-DPTC: A non-negative temporal dimension
preserved tensor completion model for missing traffic data imputation. Information Sciences, 653, 119797.
Advance online publication. 10.1016/j.ins.2023.119797
Chen, S. M., & Lee, L. W. (2010). Fuzzy multiple attributes group decision-making based on the ranking values
and the arithmetic operations of interval type-2 fuzzy sets. Expert Systems with Applications, 37(1), 824–833.
10.1016/j.eswa.2009.06.094
Dempster, A. P. (1967). Upper and lower probabilities induced by a multivalued mapping. Annals of Mathematical
Statistics, 38(2), 325–339. 10.1214/aoms/1177698950
Deng, Y. (2020). Uncertainty measure in evidence theory. Science China. Information Sciences, 63(11), 210201.
Advance online publication. 10.1007/s11432-020-3006-9
Feng, B., Cheng, F., Liu, Y., Chang, X., Wang, X., & Jin, D. (2024). Community detection on social networks
with sentimental interaction. [IJSWIS]. International Journal on Semantic Web and Information Systems, 20(1),
1–23. 10.4018/IJSWIS.341232
Gupta, B. B., & Panigrahi, P. K. (2022). Analysis of the role of global information management in advanced
decision support systems (DSS) for sustainable development. [JGIM]. Journal of Global Information Management,
31(2), 1–13. 10.4018/JGIM.320185
Hafezalkotob, A., Hafezalkotob, A., Liao, H., & Herrera, F. (2019). An overview of MULTIMOORA for
multi-criteria decision-making: Theory, developments, applications, and challenges. Information Fusion, 51,
145–177. 10.1016/j.inffus.2018.12.002
Haque, R., Islam, N., Tasneem, M., & Das, A. K. (2023). Multi-class sentiment classification on Bengali social
media comments using machine learning. International Journal of Cognitive Computing in Engineering, 4,
21–35. 10.1016/j.ijcce.2023.01.001
He, Z., Lin, M., Xu, Z., Yao, Z., Chen, H., Alhudhaif, A., & Alenezi, F. (2022). Deconv-transformer (DecT):
A histopathological image classification model for breast cancer based on color deconvolution and transformer
architecture. Information Sciences, 608, 1093–1112. 10.1016/j.ins.2022.06.091
Hu, J., Zhang, Y., Chen, X., & Liu, Y. (2013). Multi-criteria decision making method based on possibility
degree of interval type-2 fuzzy number. Knowledge-Based Systems, 43, 21–29. 10.1016/j.knosys.2012.11.007
Hu, Y., & Man, Y. (2023). Energy consumption and carbon emissions forecasting for industrial processes:
Status, challenges and perspectives. Renewable & Sustainable Energy Reviews, 182, 113405. Advance online
publication. 10.1016/j.rser.2023.113405

29
International Journal on Semantic Web and Information Systems
Volume 20 • Issue 1 • January-December 2024

Jin,
F., Zhao, Y., Zheng, X., & Zhou, L. (2023). Supplier selection through interval type-2 trapezoidal fuzzy
multi-attribute group decision-making method with logarithmic information measures. Engineering Applications
of Artificial Intelligence, 126, 107006. Advance online publication. 10.1016/j.engappai.2023.107006
Kalinke, F., Bielski, P., Singh, S., Fouché, E., & Böhm, K. (2021). An evaluation of NILM approaches on
industrial energy-consumption data. Proceedings of the Twelfth ACM International Conference on Future Energy
Systems, 239–243. 10.1145/3447555.3464863
Liu, J., Li, X., Zhang, Y., Li, T., Wei, L., & Wu, W. (2024). Failure risk assessment by multi-state dynamic
Bayesian network based on interval type-2 fuzzy sets and leaky-weighted sum algorithm: A case study of
crude oil pipelines. Expert Systems with Applications, 250, 123942. Advance online publication. 10.1016/j.
eswa.2024.123942
Liu, P., & Li, Y. (2019). An extended MULTIMOORA method for probabilistic linguistic multi-criteria group
decision-making based on prospect theory. Computers & Industrial Engineering, 136, 528–545. 10.1016/j.
cie.2019.07.052
Mendel, J. M., John, R. I., & Liu, F. (2006). Interval type-2 fuzzy logic systems made simple. IEEE Transactions
on Fuzzy Systems, 14(6), 808–821. 10.1109/TFUZZ.2006.879986
Nahar, K. M., Banikhalaf, M., Ibrahim, F., Abual-Rub, M., Almomani, A., & Gupta, B. B. (2023). A rule-based
expert advisory system for restaurants using machine learning and knowledge-based systems techniques. [IJSWIS].
International Journal on Semantic Web and Information Systems, 19(1), 1–25. 10.4018/IJSWIS.333064
Patil, Y., & Singh, P. (2024). Factors impacting adoption of climate change mitigation strategies in the Indian
iron and steel industry. In Shukla, K., Patil, Y., Estoque, R., & López de Haro, P. (Eds.), Quality of life and
climate change: Impacts, sustainable adaptation, and social-ecological resilience (pp. 233–256). IGI Global.,
10.4018/978-1-6684-9863-7.ch011
Pettinger, R., Gupta, B. B., Roja, A., & Cozmiuc, D. (Eds.). (2023). Handbook of research on digitalization
solutions for social and economic needs. IGI Global., 10.4018/978-1-6684-4102-2
Piper, J., & Rodger, J. A. (2022). Longitudinal study of a website for assessing American Presidential candidates
and decision making of potential election irregularities detection. [IJSWIS]. International Journal on Semantic
Web and Information Systems, 18(1), 1–20. 10.4018/IJSWIS.305802
Qin, J., Li, M., & Liang, Y. (2022). Minimum cost consensus model for CRP-driven preference optimization
analysis in large-scale group decision making using Louvain algorithm. Information Fusion, 80, 121–136.
10.1016/j.inffus.2021.11.001
Qin, J., & Ma, X. (2022). An IT2FS-PT3 based emergency response plan evaluation with MULTIMOORA
method in group decision making. Applied Soft Computing, 122, 108812. Advance online publication. 10.1016/j.
asoc.2022.10881235431708
Rao, M. G., Pawar, S., Priyanka, H., & Kumar Reddy, H. (2023). Minimize the energy consumption for
communication protocol in IoT. In R. Gatti, C. Singh, S. P., & S. Bhat (Eds.), Energy Systems Design for
Low-Power Computing (pp. 214–234). IGI Global. 10.4018/978-1-6684-4974-5.ch011
Ren, R., Fang, J., Hu, J., Ma, X., & Li, X. (2023). Risk assessment modeling of urban railway investment and
financing based on improved SVM model for advanced intelligent systems. [IJSWIS]. International Journal on
Semantic Web and Information Systems, 19(1), 1–19. 10.4018/IJSWIS.331596
Rosenberg, R. P., Benca, R., Doghramji, P., & Roth, T. (2023). A 2023 update on managing insomnia in primary
care: Insights from an expert consensus group. The Primary Care Companion for CNS Disorders, 25(1), Article
22nr03385. 10.4088/PCC.22nr03385
Sangaiah, A. K., Rezaei, S., Javadpour, A., & Zhang, W. (2023). Explainable AI in big data intelligence of
community detection for digitalization e-healthcare services. Applied Soft Computing, 136, 110119. Advance
online publication. 10.1016/j.asoc.2023.110119
Sangaiah, A. K., Wang, X., Lin, Y. B., Jianwei, N., & Yuan, X. (2023). Guest editorial cognitive cyber-physical
systems with AI based solutions in medical informatics. IEEE Journal of Biomedical and Health Informatics,
27(2), 586–587. 10.1109/JBHI.2023.3234603

30
 International Journal on Semantic Web and Information Systems
Volume 20 • Issue 1 • January-December 2024

Sathishkumar, V. E., Shin, C., & Yongyun, C. (2023). Steel industry energy consumption [Data set]. UCI Machine
Learning Repository. 10.24432/C52G8C
Shafer, G. (1978). A mathematical theory of evidence. Technometrics, 20(1), 106. 10.1080/00401706.1978.10489628
Tan, J., Zhou, W., Lin, L., & Jumahong, H. (2024). A review of semantic medical image segmentation based on
different paradigms. [IJSWIS]. International Journal on Semantic Web and Information Systems, 20(1), 1–25.
10.4018/IJSWIS.345246
Touqeer, M., Hafeez, A., & Arshad, M. (2020). Multi-attribute decision making using grey relational projection
method based on interval type-2 trapezoidal fuzzy numbers. Journal of Intelligent & Fuzzy Systems, 38(5),
5979–5986. 10.3233/JIFS-179684
Wang, C., Song, Z., & Fan, H. (2024). Novel evidence theory-based reliability analysis of functionally graded
plate considering thermal stress behavior. Aerospace Science and Technology, 146, 108936. Advance online
publication. 10.1016/j.ast.2024.108936
Wang, P., Fu, Y., Liu, P., Zhu, B., Wang, F., & Pamucar, D. (2024). Evaluation of ecological governance in the
Yellow River basin based on Uninorm combination weight and MULTIMOORA-Borda method. Expert Systems
with Applications, 235, 121227. Advance online publication. 10.1016/j.eswa.2023.121227
Wu, Q., Ren, H., Shi, S., Fang, C., Wan, S., & Li, Q. (2023). Analysis and prediction of industrial energy
consumption behavior based on big data and artificial intelligence. Energy Reports, 9(Suppl. 3), 395–402.
10.1016/j.egyr.2023.01.007
Xia, C., Wang, Z., Xu, G., & Dong, X. (2022). Quantitative decomposition of China’s industrial energy
consumption structure based on production-theoretical decomposition method. Journal of Cleaner Production,
379, 134467. Advance online publication. 10.1016/j.jclepro.2022.134467
Xiao, F., Cao, Z., & Jolfaei, A. (2021). A novel conflict measurement in decision-making and its application in
fault diagnosis. IEEE Transactions on Fuzzy Systems, 29(1), 186–197. 10.1109/TFUZZ.2020.3002431
Xiao, F., Wen, J., & Pedrycz, W. (2022). Generalized divergence-based decision making method with an
application to pattern classification. IEEE Transactions on Knowledge and Data Engineering, 35(7), 6941–6956.
10.1109/TKDE.2022.3177896
Xu, X., Lin, M., Luo, X., & Xu, Z. (2023). HRST-LR: A Hessian regularization spatio-temporal low rank algorithm
for traffic data imputation. IEEE Transactions on Intelligent Transportation Systems, 24(10), 11001–11017.
10.1109/TITS.2023.3279321
Zadeh, L. A. (1965). Fuzzy sets. Information and Control, 8(3), 338–353. 10.1016/S0019-9958(65)90241-X
Zadeh, L. A. (1975). The concept of a linguistic variable and its application to approximate reasoning-I.
Information Sciences, 8(3), 199–249. 10.1016/0020-0255(75)90036-5
Zhang, C., Wang, B., Li, W., & Li, D. (2024). Incorporating artificial intelligence in detecting crop diseases:
Agricultural decision-making based on group consensus model with MULTIMOORA and evidence theory.
Crop Protection (Guildford, Surrey), 179, 106632. Advance online publication. 10.1016/j.cropro.2024.106632
Zhang, H., Dong, Y., Chiclana, F., & Yu, S. (2019). Consensus efficiency in group decision making: A
comprehensive comparative study and its optimal design. European Journal of Operational Research, 275(2),
580–598. 10.1016/j.ejor.2018.11.052
Zhang, J., Lin, M., Pan, Y., & Xu, Z. (2023). CRFTL: Cache reallocation-based page-level flash translation layer
for smartphones. IEEE Transactions on Consumer Electronics, 69(3), 671–679. 10.1109/TCE.2023.3264217
Zhang, R., Xu, Z., & Gou, X. (2021). An integrated method for multi-criteria decision-making based on the
best-worst method and Dempster-Shafer evidence theory under double hierarchy hesitant fuzzy linguistic
environment. Applied Intelligence, 51(2), 713–735. 10.1007/s10489-020-01777-2
Zhang, S., & Xiao, F. (2023). A TFN-based uncertainty modeling method in complex evidence theory for
decision making. Information Sciences, 619, 193–207. 10.1016/j.ins.2022.11.014
Zhao, X., & Zhu, J. (2022). Industrial restructuring, energy consumption and economic growth: Evidence from
China. Journal of Cleaner Production, 335, 130242. Advance online publication. 10.1016/j.jclepro.2021.130242

31
International Journal on Semantic Web and Information Systems
Volume 20 • Issue 1 • January-December 2024

Zheng, Q., Liu, X., Wang, W., Wu, Q., Deveci, M., & Pamucar, D. (2023). The integrated prospect theory with
consensus model for risk analysis of human error factors in the clinical use of medical devices. Expert Systems
with Applications, 217, 119507. Advance online publication. 10.1016/j.eswa.2023.119507
Zhou, L., Wang, Y., & Jiang, Y. (2018). Investment project assessment by a MAGDM method based on the ranking
of interval type-2 fuzzy sets. Journal of Intelligent & Fuzzy Systems, 35(2), 1875–1888. 10.3233/JIFS-171403
Zhou, Q., Ye, C., & Geng, X. (2024). A decision framework of offshore wind power station site selection using
a MULTIMOORA method under Pythagorean hesitant fuzzy environment. Ocean Engineering, 291, 116416.
Advance online publication. 10.1016/j.oceaneng.2023.116416

32
 International Journal on Semantic Web and Information Systems
Volume 20 • Issue 1 • January-December 2024

Bingjie Wang is a dedicated graduate student with a passion for life and a positive outlook. She is currently pursuing
her PhD's degree at Shanxi University, where she consistently demonstrates a strong commitment to academic
excellence. With a proactive approach to learning, she actively engages in her studies and continuously strives for
personal and intellectual growth. With a strong work ethic and a genuine enthusiasm for learning, Wang is poised
to make urther contributions to her feld and looks orward to continuing her academic journey with enthusiasm
and determination.

Chao Zhang is an associate professor at the Institute of Intelligent Information Processing, Shanxi University. His
main study interests include data mining, granular computing and intelligent decision making. In recent years, he
has published more than 60 papers in journals including IEEE Transactions on Computational Social Systems, IEEE
Transactions on Fuzzy Systems, IEEE Transactions on Consumer Electronics, Information Sciences, Information
Fusion, International Journal of Approximate Reasoning, ACM Transactions on Asian and Low-Resource Language
Information Processing, Computers in Industry, and Applied Mathematical Modelling. Among them, one paper has
been selected as “ESI highly cited paper.” He has published 3 academic monographs with national publishers and
published 2 national invention patents. He has been awarded frst prize or Outstanding Achievements in Scientifc
Research in Institutions o Higher Learning in Shanxi Province, the second prize or the Outstanding Achievement
Award in Social Sciences in Shanxi Province, two Excellent Academic Paper awards in Taiyuan City, the ACM
Excellent Doctoral Dissertation Award in Taiyuan Chapter, and the best student paper award in CGCKD.

Arun Kumar Sangaiah received a PhD from the School of Computer Science and Engineering, VIT University,
Vellore, India in 2007 and 2014. He is currently a full professor with the National Yunlin University of Science and
Technology, Douliu, Taiwan. He has authored/coauthored more than 300 research articles in refereed journals,
including IEEE Transactions on Industrial Informatics, IEEE Transactions on Intelligent Transportation Systems,
IEEE Transactions on Network Science and Engineering, IEEE Transactions on Emerging Topics in Computational
Intelligence, IEEE Systems Journal, IEEE Sensors Journal, IEEE Internet of Things Journal, and ACM Transactions
on Sensor Networks. He has also published 11 edited books, as well as 1 patent (held and fled), 3 projects,
including one funded by the Ministry of IT of India, and few international projects (Chinese Academy of Sciences,
Guangdong Research Fund, Australian Research Council) worth o 500 000 USD. Dr. Sangaiah is the recipient o
many awards, such as the Clarivate Highly Cited Researcher award, the Yushan Young Scholar Fellowship, the Top
2% Scientist award, the PIFI-CAS Fellowship, the Top-10 Outstanding Researcher award, and the CSI Signifcant
Contributor award. He is also editor-in-chief and associate editor of various reputable ISI journals. He is a visiting
scientist (2018–2019) with CAS, China, and visiting researcher (2019–2020) o Université Paris-Est, France.

Mohammed J. F. Alenazi (Senior Member, IEEE) received his B.S., M.S., and PhD degrees in computer engineering
from the University of Kansas in 2010, 2012, and 2015, respectively. He is currently an Associate Professor in
computer engineering with King Saud University. His research interests include, but are not limited to, the design,
implementation, and analysis of resilient and survivable networks; network routing design and implementation; the
development and simulation of network architectures and protocols; the performance evaluation of communication
networks; an algorithmic graph approach or modeling networks; and mobile ad hoc network (MANET) routing
protocols. He is a member of the ACM.

Salman A. AlQahtani is currently a full professor with the Department of Computer Engineering in the College of
Computer and Information Sciences, King Saud University, Riyadh, Saudi Arabia. His main research interests include
radio resource management for wireless and cellular networks (4G, 5G, the IoT, Industry 4.0, LTE, LTE-advanced,
emtocell, cognitive radio, and cyber sovereignty) with a ocus on call admission control, packet scheduling, radio
resource sharing, and the quality of service guarantees for data services.

K.S Sendhil Kumar works as an associate proessor (Senior) in the School o Computer Science and Engineering
at VIT, Vellore. He received his PhD in computer science and engineering rom VIT, Vellore, his ME (CSE) rom
Anna University, and his BE (CSE) rom University o Madras. He has 17 years o experience in teaching and
research. He has published many papers in international and national journals and presented at conferences on
cloud computing, machine learning, IoT, and big data. He acts as a reviewer for many international journals and
conferences. He is a life member of life member of Computer Society of India. His research interests include cloud
computing, block chain, IoT and machine learning.

33