An interval-valued atanassov’s intuitionistic fuzzy multi-attribute group decision making method based on the best representation of the WA and OWA operators

Corresponding Author: bedregal@dimap.ufrn.br
http://dx.doi.org/10.22105/jfea.2021.306164.1162
E-ISSN: 2717-3453 | P-ISSN: 2783-1442
|
Abstract
1 | Introduction
From the seminal paper [71] on fuzzy set theory, several extensions for this theory have been
proposed [18]. Among them, we stress “Interval-valued Fuzzy Sets Theory” [10], [19], [72] and
“Atanassov’s Intuitionistic Fuzzy Sets Theory” [2], [5], [25], [26]. Although they are mathematically
equivalents, they capture different kinds of uncertainty in the membership degrees, i.e. they have
different semantics [61]. The first one takes in account the intrinsic difficulty to determine the exact
membership degree of an object to some linguistic term; in this case, an expert provides an interval
which expresses his uncertainty on such degree. The second one adds an extra degree to the usual
fuzzy sets in order to model the hesitation and uncertainty about the membership degree. In fuzzy
set theory, the non-membership degree is by default the complement of the membership degree, i.e.
Journal of Fuzzy Extension and Applications
www.journal-fea.com
J. Fuzzy. Ext. Appl. Vol. 2, No. 3 (2021) 239–261.
Paper Type: Research Paper
An Interval-Valued Atanassov’s Intuitionistic Fuzzy Multi-
attribute Group Decision Making Method Based on the Best
Representation of the WA and OWA Operators
Ivanosca Andrade Da Silva1, Berta Bedregal2, Benjamın Bedregal3,* , Regivan Hugo Nunes Santiago3
1 Directorate of Material and Heritage, Federal University of Rio Grande do Norte, Natal, RN, Brazil; ivanosca@yahoo.com.br.
2 Production Engineering Course state University of Rio de Janeiro; leticia.berta@gmail.com.
3 Department of Informatics and Applied Mathematics, Federal University of Rio Grande do Norte, Natal, RN, Brazil;
bedregal@dimap.ufrn.br; regivan@dimap.ufrn.br.
Citation:
Da Silva1, I. A., Bedregal, B., Bedregal, B., & Santiago, R. H. N. (2021). An interval-valued
atanassov’s intuitionistic fuzzy multi-attribute group decision making method based on the best
representation of the WA and OWA operators. Journal of fuzzy extension and application, 2 (3), 239-
261.
Accept: 19/09/2021
Revised: 09/09/2021
Reviewed: 02/08/2021
Received: 04/06/2021
In this paper we extend the notion of interval representation for interval-valued Atanassov’s intuitionistic representations,
in short Lx-representations, and use this notion to obtain the best possible one, of the Weighted Average (WA) and
Ordered Weighted Average (OWA) operators. A main characteristic of this extension is that when applied to diagonal
elements, i.e. fuzzy degrees, they provide the same results as the WA and OWA operators, respectively. Moreover, they
preserve the main algebraic properties of the WA and OWA operators. A new total order for interval-valued Atanassov’s
intuitionistic fuzzy degrees is also introduced in this paper which is used jointly with the best Lx-representation of the
WA and OWA, in a method for multi-attribute group decision making where the assesses of the experts, in order to take
in consideration uncertainty and hesitation, are interval-valued Atanassov’s intuitionistic fuzzy degrees. A characteristic
of this method is that it works with interval-valued Atanassov’s intuitionistic fuzzy values in every moments, and
therefore considers the uncertainty on the membership and non-membership in all steps of the decision making. We
apply this method in two illustrative examples and compare our result with other methods.
Keywords: Interval-Valued Atanassov’s intuitionistic fuzzy sets, WA and OWA operators, Lx-representations, Total
orders, Multi-Attribute group decision making.
Licensee Journal
of Fuzzy Extension and
Applications. This article
is an open access article
distributed under the
terms and conditions of
the Creative Commons
Attribution (CC BY)
license
(http://creativecommons.
org/licenses/by/4.0).

240
Da
Silva
et
al.
|J.
Fuzzy.
Ext.
Appl.
2(3)
(2021)
239-261
1- µA (x) , meaning that there is no doubt or hesitation in the membership degree. In [3], both extensions
are mixed by considering that we can also have an uncertainty or imprecision in the membership and
non-membership degrees if we model them with intervals. This results in other extension of fuzzy set
theory, known as Interval-Valued Atanassov’s Intuitionistic Fuzzy Sets (IVAIFS). Several applications
of IVAIFS, and extensions of usual fuzzy notions to the IVAIFS framework have been made, see for
example [4], [7], [21], [32], [51], [64].
Besides, Group Decision Making (GDM) and Multi-attribute Decision Making (MADM) are the most
well know branches of decision making. GDM consists in choosing of one or more alternatives among
several ones by a group of decision makers (experts), probably with a weight of confidence [24]. MADM
choosing one or more alternatives among several ones based in the assesses of an expert his opinion of
how much the alternative fulfills a criteria or satisfies an attribute. Usually, a weighting vector for the
attributes is associated, in order to represent the importance of an attribute in the overall decision
problem. Nevertheless, complex decision making problems usually need to consider a group of experts
as well as a set of criteria or attributes, i.e. a Multi-attribute Group Decision Making (MAGDM) [28],
[43], [55], [59].
Fuzzy logic, by their nature, has played an important role in the field of decision making, since decision
makers can be subject to uncertainty expressed in terms of fuzzy degrees [46], [47], [55], [57]. An
important mathematical tool for fuzzy decision-making are Weighted Average (WA) and the Ordered
Weighted Average (OWA) operators introduced in [69], which have triggered their “extension” for
Interval-Valued Atanassov’s Intuitionistic Fuzzy Values (IVAIFV) – see for example [65], [67].
However, in the cited cases, the proposed interval-valued Atanassov’s intuitionistic OWA, although of
preserve some algebraic properties of the OWA (monotonicity, idempotency, symmetry and
boundedness [16]), have not the same behaviour as the OWA when applied to diagonals elements.
In [11], [54], in order to formalize the principle of correctness of interval computation [37], it was
introduce the notion of interval representation of real functions. In addition, the best of the interval
representations of a real function models the notion of optimality in interval computing. This notion
has been used in the context of interval-valued fuzzy functions, to obtain interval-valued t-norms (t-
conorms, overlap functions, fuzzy negations and fuzzy implications) from t- norms (t-conorms, overlap
functions, fuzzy negations and fuzzy implications) in [1], [8], [14], [34]. In this paper we extend the
notion of interval-valued representation and the best interval-valued representation of fuzzy functions
for the interval-valued Atanassov’s intuitionistic representations of fuzzy and interval- valued fuzzy
functions. In particular, we provide a novel extension of the WA and OWA operator for IVAIFS, based
on the best interval-valued Atanassov’s intuitionistic fuzzy representation, which preserve the main
properties of the OWA operators and when restrict to the diagonals elements it is an OWA in 0, 1 . This
new IVAIFAF OWA together with some total orders for IVAIFV are used to develop a method to rank
alternatives from the individual interval-valued Atanassov’s intuitionistic decision matrices of a group
of experts reflecting how much each alternative satisfy each attribute. Two illustrative examples are
considered in order to show the use of the method and to show that the final ranking of alternatives
obtained by the method is adequate.
This paper is organized as follows: Section 2 introduces Atanassov intuitionisc and interval-valued fuzzy
sets, the score and accuracy index and the notion of representation in particular in the interval-valued
and Atanassov intuitionisc best representation of the WA and OWA operators. In Section 3 it is consider
the notion of interval-valued intuitionistic fuzzy set and some orders for interval-valued. Atanassov’s
intuitionistic fuzzy values are presented. In particular, based in a novel notion of membership and
subsets, interval-valued intuitionistic fuzzy degrees are seen as an interval of interval-valued fuzzy
degrees and based in this
point of view a new total order for IVAIFV is provided. In Section 4 it is introduced the notion of
IVAIFV representation and it is provide a canonical way of obtain the best representation of an interval-

241
An
interval-valued
atanassov's
intuitionistic
fuzzy
multi-attribute
group
decision
making
method
based
on
the
best
representation
of
the
WA
and
OWA
operators
valued fuzzy function and of a fuzzy function, which is used to obtain the best IVAIFV representation
of the WA and OWA operators. In Section 5 the total orders on IVAIFV and the best IVAIFV
representation of the WA and OWA are used to develop a method to solve MAGDMP and this method
is used in two illustrative examples. Finally in Section 6 some final remarks on the paper are provided.
2| Preliminaries
Atanassov in [2] extended the notion of fuzzy sets, by adding an extra degree to model the hesitation or
uncertainty in the membership degree. This second degree is called non-membership degree. In fuzzy set
theory, by default, this non- membership degree is given by the complement of the membership degree,
i.e. one minus the membership degree, and therefore is fixed whereas in Atanassov intuitionistic fuzzy sets
the non-membership degree may take any value between zero and one minus the membership degree.
Definition 1. [2]. Let X be a non-empty set and two functions A
μ , A
ν : X [0 ,1]
 . Then
is an Atanassov Intuitionistic Fuzzy Set (AIFS) over X if A A
μ ( x) ν ( x) 1
  for each x X
 .
The functions A
μ and A
ν provide the membership and non-membership degrees of elements in X to the
AIFS A. Let 2
L {(x, y) [0,1] / x y 1}

    . Elements of L
are calledL
-values. We define the
projections l, r :L [0,1]

 by l(x,y)=x and r(x,y)=y, but by notational simplicity, we will denote x
%
and x
%
instead of l(x) and r(x), respectively.
The usual partial order on L
is the following:
Deschrijver and Kerre [33] proved that L
L , 

   is a complete lattice and therefore that AIFS are a
particular kind of L-fuzzy set, in the sense of Goguen [35].
Let A be an AIFS over X. The intuitionistic fuzzy index1 of an element x X
 to A is given by
*
A A A
π ( x) 1 -μ ( x) ν ( x)
  . In particular, the intuitionistic fuzzy index of *
x L
 is defined in a similar way,
i.e. *
A
π ( x) 1 -l( x) r( x)
  . This index measures the hesitation degree in each *
x L
 .
In [27], Chen and Tan, introduce the notion of score of a L
-value as the function *
S : L [ 1,1]

  defined
by
In [38], Hong and Choi, introduce the notion of accuracy function for an L
-value as the function
*
h : L [0,1]

 defined by
1
In the seminal paper on AIFS, i.e. in [2], this index was called degree of indeterminacy of an element 
x X to A
A A
A {(x,μ (x), ν (x)) / x X},
 
  
*
x L y if x y and y x
%
%
% %
.
 
*
S (x) x x
%
%
(1)
 
*
h (x) x x
%
%
(2)

242
Da
Silva
et
al.
|J.
Fuzzy.
Ext.
Appl.
2(3)
(2021)
239-261
Xu and Yager in [68], based on the score and accuracy index on L
and with the goal of rank L
-
values, introduce the total order on L
defined by
In [36], [40], [52], [72] and in an independent way, fuzzy set theory was extended by considering
subintervals of the unit interval [0,1] instead of a single value in [0,1]. The main goal was to represent
the uncertainty in the process of assigning the membership degrees.
Definition 2. Let X be a non-empty set and L {[a,b] / 0 a b 1}
    be the set of closed
subintervals of [0,1]. An Interval-Valued Fuzzy Set (IVFS) A over X is an expression
Where A
μ : X L.

Define the projections1 , : L [0,1]
   by ([a,b]) = a and ([a,b]) b.
  
For notational simplicity, for an arbitrary X L
 , we will denote (x)
 and (x)
 by X and X ,
respectively. An interval X L
 is degenerate if X = X , i.e. X=[x,x] for some x [0,1]
 . Given X L
 ,
we denote its standard complement [1 X , 1-X]
 by X. A more general notion of complement (or
negation) for L
-values can be found in [8].
We can consider the following partial order on L,
As it is well-known, L
L,
   is a complete lattice and so it can be seen as a Goguen L-fuzzy set.
As pointed by Moore in [45], an interval has a dual nature: as a set of real numbers and as a new kind of
number (an ordered pair of real numbers with the restriction that the first component is smaller than or
equal to the second one). The order L
 is an order which stresses the nature of ordered pair for elements
in L whereas the inclusion of sets stresses the nature of set for elements in L. Nevertheless, the inclusion
order on L can also be expressed using the ordered pair nature as follows:
The score and accuracy function for interval fuzzy values, i.e. of an arbitrary interval X L
 are defined
as follows:
Where v( X ) X X
  and w(X) X X
  .
1
These projections are particular cases of Atanassov’s Kα-operator for intervals [19], [48].
   
* * * * * *
YY
x y if s (x) s (y) or (s (x) s (y) and h (x) h (y)) . (3)
 
A
A {(x,μ (x)) / x X}.
L
X Y iff X Y and X Y.
  
   
X Y iff Y X X Y.
s(X) = v(X) -1 and h(X)= 1-w(x).

243
An
interval-valued
atanassov's
intuitionistic
fuzzy
multi-attribute
group
decision
making
method
based
on
the
best
representation
of
the
WA
and
OWA
operators
As it is well known, the lattices L
L , 

   and L
L,
   are isomorphic. The map *
ρ : L L
 defined by
ρ( X ) : (X , 1-X) is a such an isomorphism. Although both lattices are algebraically equivalent, from a
semantical point of view they are different [61].
Remark 1. Note that, the score and the accuracy indexes on L and L
are related as follows: *
s s ο ρ

and *
h h ο ρ
 . Notice that the partial order XY
 on L obtained from the partial order XY
 in Eq.(3) by
using this isomorphism, i.e. XY
X Y
 iff XY
ρ( X ) ρ(Y )
 , can be equivalently obtained as following:
Bustince et al. [22] introduced the notion of admissible orders in the context of interval-valued fuzzy
functions in order to always be possible to compare intervals which is important in some kind of
applications [23]. An order  on L is admissible if it refines L L
  , i.e. X Y
 whenever L
X Y
 . In
particular XY
 is an admissible order. Other examples of admissible orders can be found in [53]. In addition,
when we translate the notion of intuitionistic fuzzy index for interval values, we get the interval-valued
fuzzy index *
Π(X) π (ρ(X)) X X w(X)
    for eachX L
 . Thus, the length of an interval is a measure
of their indeterminacy or imprecision.
2.1| The Best L and L* Representation of the OWA Operator
In [13], it was adapted the notion of interval representation of [11], [54] in the context of interval-valued
fuzzy sets theory for the particular case of the intervalvalued t-norms. Interval representation captures, in
a formal way, the property of correctness of interval functions in the sense of [37]. From then, interval
representations of several other connectives and fuzzy constructions (see for example [8], [12], [49]) have
been studied. Here we are interested in considering the case of n-ary increasing fuzzy functions. Let’s start
recalling some notions.
Definition 3. Let n
f : [0,1] [0,1]
 be an n-ary function. A function n
F : L L
 is an interval
representation or L-representation of f if for each 1 n
X ,.....,X L
 and i i
x X
 with i=1,....,n we have that
1 n 1 n
f( x ,.....,x ( X ,...
F )
..,X
)  .
Let n
F ,G : L L
 . We write L
F G
  , if for any 1 n
X ,.....,X L
 , 1 n 1 n
G( X ,.....,X F( X ,... )
..,X
)  .
Notice that if X ,Y L
 and X Y
 then )
h(X) h(Y
 . Thus, L
F G
  means that G is always more
accurate than F, i.e. 1 n 1 n
h )) h(
(F( X ,.....,X G( X ,.....,X ))
 for any 1 n
X ,.....,X L
 . Notice also that if G is
an L-representation of a function f and L
F G
  then F is also an L-representation of f, but less accurate
than G. Therefore, G is a better L-representation of f than F.
Proposition 1. [34]. Let n
f : [0,1] [0,1]
 be an n-ary increasing fuzzy function. Then the function
n
f̂ : L L
 defined by
is an L-representation of f. Moreover, for any other L-representation F of f, L
ˆ
F f .
f̂ is therefore the more accurate L-representation of f, i.e. the best L-representation w.r.t. the L
order. So
f̂ has the property of optimality in the sense of [37].
  
XY XY
X Y iff s(X) s(Y) or (s(X) s(Y) and h(X) h(Y)).
p (4)

1 n 1 n 1 n
f̂(X ,....,X ) [f(X ,....,X ),f(X ,....,X )] , ( )

244
Da
Silva
et
al.
|J.
Fuzzy.
Ext.
Appl.
2(3)
(2021)
239-261
Remark 2. [10]. An important characteristic of the best L-representation is that when we identify points
and degenerate intervals, via the merging 
m(x) [x,x] , f and f̂ have the same behavior, i.e.

1 n 1 n
ˆ
m(f( x ,...,x )) f(m( x ),...,m( x )). Another property of the best L-representation of some increasing
function is that it is isotone with respect to both, the inclusion order and the  L order, i.e. if 
i i
X ,Y L
and 
i 1,...., n then 𝑓̂(𝑋1, … , 𝑋𝑛) ⊆ 𝑓̂(𝑌1, … , 𝑌𝑛) and, analogously, if 𝑋𝑖 ≤𝕃 𝑌𝑖 for each 𝑖 = 1,… , 𝑛 then
𝑓̂(𝑋1, … , 𝑋𝑛) ≤𝕃 𝑓̂(𝑌1, … , 𝑌𝑛).
Let Λ = (𝜆1, … , 𝜆𝑛) ∈ [0,1]𝑛
be an n-ary weighting vector, i.e. ∑𝑛
𝑖=1 𝜆𝑖 = 1. The weighted average (WA)
operator is defined by
The Ordered Weighted Averaging (OWA) operator introduced by Yager [69] is defined by
Where 𝜎: {1,… , 𝑛} → {1,… , 𝑛} is the permutation such that 𝑥𝜎(𝑖) ≥ 𝑥𝜎(𝑖+1) for any 𝑖 = 1, … , 𝑛 − 1, i.e. it
orders in decreasing way a n-tuple of values in [0,1] and so 𝑥𝜎(𝑖) is the 𝑖th greatest element of {𝑥1, … , 𝑥𝑛}.
Notice that,
Several interval-valued and Atanassov intuitionistic extensions of the OWA operator have been
proposed (see for example [15], [44], [70]), but most of them are not 𝕃 (𝐿∗
)-representations of the OWA
operator and do not reduce to the fuzzy OWA operator when applied to degenerate intervals.
The best 𝕃-representation of 𝑜𝑤𝑎Λ is the interval-valued function 𝑜𝑤𝑎𝛬
̂:𝕃𝑛
→ 𝕃 defined by
Where 𝑋𝜏𝑖 = [𝑋𝜏1(𝑖),𝑋𝜏2(𝑖)]; 𝜏1, 𝜏2: {1, … , 𝑛} → {1, … , 𝑛} are permutations such that 𝑋𝜏1(𝑖) ≥ 𝑋𝜏1(𝑖+1) and
𝑋𝜏2(𝑖) ≥ 𝑋𝜏2(𝑖+1) for any 𝑖 = 1, … , 𝑛 − 1; the scalar product is the usual in interval mathematics (see [45]),
i.e. for any 𝜆 ∈ [0,1] and 𝑋, 𝑌 ∈ 𝕃, 𝜆𝑋 = [𝜆𝑋,𝜆𝑋] and the sum is w.r.t. the limited addition defined by
𝑋[+]𝑌 = [min(𝑋 + 𝑌, 1), min(𝑋 + 𝑌, 1)]. Notice that, in this case, because∑𝑛
𝑖=1 𝜆𝑖 = 1,
Where [∑𝑛
𝑖=1 ] is the sommatory with respect to [+] and ∑𝑛
𝑖=1 is the sommatory with respect the usual
addition between intervals (see [45]).
Note that for each term in the sum above, lower and upper bounds from different intervals may be
considered for a given weight𝜆𝑖. For example, for 𝜆1 = 0.2, 𝜆2 = 0.3, 𝜆3 = 0.5, 𝑋1 = [0.6,0.8], 𝑋2 =
[0.7,0.9] and 𝑋3 = [0.5,1] we have that [∑3
𝑖=1 ]𝜆𝑖𝑋𝑖 = [𝑚𝑖𝑛(0.2 ⋅ 0.6 + 0.3 ⋅ 0.7 + 0.5 ⋅ 0.5,1), 𝑚𝑖𝑛(0.2 ⋅ 0.8 +
0.3 ⋅ 0.9 + 0.5 ⋅ 1,1)] = [0.58,0.93] = [0.2 ⋅ 0.6 + 0.3 ⋅ 0.7 + 0.5 ⋅ 0.5,0.2 ⋅ 0.8 + 0.3 ⋅ 0.9 + 0.5 ⋅ 1] =
∑3
𝑖=1 𝜆𝑖𝑋𝑖.
waΛ(x1, … , xn) = ∑
n
i=1
λixi.
owaΛ(x1, … , xn) = ∑
n
i=1
λixσ(i).
owaΛ(x1, … , xn) = waΛ(xσ(1), … , xσ(n)). (6)
owaΛ
̂ (X1, …, Xn) = [owaΛ(X1, …, Xn), owaΛ(X1, … , Xn)] = ∑
n
i=1
λiXτ(i).
[∑
n
i=1
] λiXi = [min(∑
n
i=1
λiXi,1) , min (∑
n
i=1
λiXi,1)] = [∑
n
i=1
λiXi,∑
n
i=1
λiXi]
= ∑
n
i=1
λiXi.
(7)

245
An
interval-valued
atanassov's
intuitionistic
fuzzy
multi-attribute
group
decision
making
method
based
on
the
best
representation
of
the
WA
and
OWA
operators
Analogously, a function 𝐹: (𝐿∗
)𝑛
→ 𝐿∗
is an 𝐿∗
-representation of a function 𝑓: [0,1]𝑛
→ [0,1] if for each 𝐱𝑖 ∈
𝐿∗
and 𝑥𝑖 ∈ [𝐱𝑖,1 − 𝐱𝑖
̃], with 𝑖 = 1, … , 𝑛,
Let 𝐹, 𝐺: (𝐿∗
)𝑛
→ 𝐿∗
. We denote by 𝐹 ⊑𝐿∗ 𝐺, if for any x1, … , x𝑛 ∈ 𝐿∗
, 𝐺(x1, … , x𝑛) ⊆𝐿∗ 𝐹(x1, … , x𝑛), where
𝑥 ⊆𝐿∗ 𝑦 if 𝑥 ≤ 𝑦 and 𝑥̃ ≤ 𝑦̃. Notice that although of this order be the usual on 𝑅2
, considering the
mathematical equivalence of 𝐿∗
and 𝕃, we have that 𝑥 ⊆𝐿∗ 𝑦 iff 𝜌−1
(𝑋) ⊆ 𝜌−1
(𝑌). Thus, 𝐹 ⊑𝐿∗ 𝐺 means than
the result of 𝐺 is always more accurate than the result of 𝐹, i.e. ℎ∗
(𝐹(𝑥1, … , 𝑥𝑛)) ≤ ℎ∗
(𝐺(𝑥1, … , 𝑥𝑛)) for any
𝑥1, … , 𝑥𝑛 ∈ 𝐿∗
.
Proposition 2. Let 𝑓: [0,1]𝑛
→ [0,1] be an increasing function. Then the function 𝑓: (𝐿∗
)𝑛
→ 𝐿∗
defined by
is the greatest 𝐿∗
-representation of 𝑓 w.r.t. ⊑𝐿∗ order and so is the best one.
Proof. If 𝑥𝑖 ∈ [𝐱𝑖,1 − 𝐱𝑖
̃] for each 𝑖 = 1, … , 𝑛, then because 𝑓 is increasing we have that 𝑓(𝑥1, … , 𝑥𝑛) ≤
𝑓(𝑥1, … , 𝑥𝑛) ≤ 𝑓(1 − 𝑥1
̃, … ,1 − 𝑥𝑛
̃) and therefore, 𝑓(𝑥1, … , 𝑥𝑛) ≤ 𝑓(𝑥1, … , 𝑥𝑛) ≤ 1 − 𝑓(𝑥1, … , 𝑥𝑛)
̃ . So,
𝑓(𝑥1, … , 𝑥𝑛) is an 𝐿∗
-representation of 𝑓.
Now, suppose that 𝐹 is another 𝐿∗
-representation of 𝑓, then by Eq. (8) and because 𝑓 is increasing, we
have that 𝐹(𝑥1, … , 𝑥𝑛) ≤ 𝑓(𝑥1, … , 𝑥𝑛) ≤ 𝑓(1 − 𝑥1
̃, … ,1 − 𝑥𝑛
̃) ≤ 1 − 𝐹(𝑥1, … , 𝑥𝑛)
̃ . Therefore,
𝑓(𝑥1, … , 𝑥𝑛) ⊆𝐿∗ 𝐹(𝑥1, … , 𝑥𝑛), i.e. 𝐹 ⊑∗
𝑓.
Moreover, if 𝑓 is an aggregation function then 𝑓 is also an 𝐿∗
-valued aggregation function [42] (Lemma 1).
Clearly, 𝑓 = 𝜌 ∘ 𝑓̂ ∘ 𝜌−1
, or equivalently, 𝑓̂ = 𝜌−1
∘ 𝑓 ∘ 𝜌. Therefore, 𝑜𝑤𝑎Λ it is the best 𝐿∗
-representation of
𝑜𝑤𝑎Λ .
Proposition 3. Let f, g: [0,1]n
→ [0,1]. If f ≤ g then f̂ ≤ g
̂ and f ≤ g.
Proof. Straightforward.
Remark 3. owa as well as owa
̂ are interval-valued and Atanassov intuitionistic aggregation functions in
the sense of [42]. Moreover, both are symmetric and idempotent, and as a consequence of the above
proposition, they are bounded by owa(0,…,0,1)
̂ (owa(0,…,0,1)), i.e. min
̂ (min) and owa(1,0,…,0)
̂ (owa(1,0,…,0)), i.e.
max
̂ (max).
3| Interval-Valued Atanassov’s Intuitionistic Fuzzy Sets
Definition 4. [3]. An IVAIFS 𝐴 over a nonempty set 𝑋 is an expression given by
where 𝜇𝐴,𝜈𝐴:𝑋 → 𝕃 with the condition 𝜇𝐴(𝑥) + 𝜈𝐴(𝑥) ≤ 1.
Deschrijver and Kerre [33] provide an alternative approach for Atanassov intuitionistic fuzzy sets in term
of L-fuzzy sets in the sense of Goguen [35]. Analogously, we can also see IVAIFS as a particular case of
L-fuzzy set by considering the complete lattice 〈𝕃∗
,≤𝕃∗〉 where
F(x1, … , xn) ≤ f(x1, … , xn) ≤ 1 − F(x1, …, xn)
̃ . (8)
f(x1, … , xn) = (f(x1, … , xn), 1 − f(1 − x1
̃, …,1 − xn
̃ )).
A = {(x, μA(x),νA(x))/x ∈ X},

246
Da
Silva
et
al.
|J.
Fuzzy.
Ext.
Appl.
2(3)
(2021)
239-261
And
Notice that 0𝕃∗ = ([0,0], [1,1]) and 1𝕃∗ = ([1,1],[0,0]). Analogously to the case of 𝐿∗
, we define the
projections 𝑙, 𝑟: 𝕃∗
→ 𝕃 by
and for each 𝑋 ∈ 𝕃∗
, we denote 𝑙(𝑋) and 𝑟(𝑋) by 𝑋 and 𝑋
̃, respectively.
Elements of 𝕃∗
will be called 𝕃∗
-values. An 𝕃∗
-value 𝑋 is a semi-diagonal element if 𝑋 and 𝑋
̃ are
degenerate intervals. 𝑋 ∈ 𝕃∗
is a diagonal element if 𝑋 + 𝑋̃ = [1,1] i.e. if 𝑋 = ([𝑥,𝑥], [1 − 𝑥, 1 − 𝑥]) for
some 𝑥 ∈ [0,1]. We denote by 𝒟𝑆 and 𝒟 the sets of semi-diagonal and diagonal elements of 𝕃∗
,
respectively. Clearly, 𝒟 ⊆ 𝒟𝑆 and there is a bijection between [0,1] and 𝒟 (𝜙(𝑥) = ([𝑥, 𝑥], [1 − 𝑥, 1 − 𝑥])),
between 𝐿∗
and 𝒟𝑆 (𝜓(𝑥) = ([𝑥,𝑥], [𝑥̃, 𝑥̃])) and between 𝕃 and 𝒟𝑆 (𝜑(𝑋) = ([𝑋, 𝑋],[𝑋, 𝑋]𝑐
), i.e. 𝜑 = 𝜓 ∘ 𝜌)
[29].
3.1 | Some indexes for 𝕃∗
-Values
In [50] the Atanassov intuitionistic fuzzy index was extended for IVAIFS, in order to provide an interval
measure of the hesitation degree in IVAIFS. Let 𝐴 be an IVAIFS over a set 𝑋. The interval-valued
Atanassov intuitionistic fuzzy index of an element 𝑥 ∈ 𝑋 for the IVAIFS 𝐴 is determined by the
expression Π∗
(𝑥) = [1,1] − 𝜇𝐴(𝑥) − 𝜈𝐴(𝑥). In an analogous way the interval-valued Atanassov
intuitionistic fuzzy index of an (𝑋, 𝑌) ∈ 𝕃∗
is defined by
The Chen and Tan score measure was extended for 𝕃∗
in [66] 1 and [41].
In this paper we consider Xu’s definition: Let 𝑆: 𝕃∗
→ [−1,1] be defined by
For each 𝑋 ∈ 𝕃∗
, 𝑆(𝑋) is called the score of 𝑋.
Remark 4. 𝑆 when applied to semi-diagonal elements is the same, up to an isomorphism 𝜓, as 𝑠∗
, i.e.
𝑆(𝜓(𝑥)) = 𝑠∗
(𝑥) for any 𝑥 ∈ 𝐿∗
. Analogously, 𝑆 when applied to semi-diagonal elements is the same, up to
an isomorphism φ, as s, i.e. S(φ(X)) = s(X) for any X ∈ 𝕃. Moreover, the range of S ([−1,1]) is the same
as that of s∗
and S can be obtained from s and s∗
, as shown by the Eq. (10).
Since we can have two different 𝕃∗
-values with the same score, for example 𝑆([0.2,0.3], [0.4,0.5]) =
𝑆([0.1,0.2],[0.3,0.4]) = −0.2, the score determines just a pre-order on 𝕃∗
:
1 Because this reference is in Chinese, we are based on the definition in [64].
𝕃∗
= {(𝑋, 𝑌) ∈ 𝕃 × 𝕃/𝑋 + 𝑌 ≤ 1}.
(X1, X2) ≤𝕃∗ (Y1, Y2) iff X1 ≤𝕃 Y1 and Y2 ≤𝕃 X2.
l(X1, X2) = X1 and r(X1, X2) = X2,
Π∗
(X,Y) = [1,1] − X − Y. (9)
S(X) =
v(X) − v(X̃)
2
.
S(X) =
s∗ (s(X), s(X̃))
2
.
(10)
X ≤S Y iff S(X) ≤ S(Y).

247
An
interval-valued
atanassov's
intuitionistic
fuzzy
multi-attribute
group
decision
making
method
based
on
the
best
representation
of
the
WA
and
OWA
operators
Since ≤𝑆 is a pre-order, it defines the following natural equivalence relation: 𝑋 ≡𝑆 𝑌 iff 𝑋 ≤𝑆 𝑌 and 𝑌 ≤𝑆 𝑋
Another important index for 𝕃∗
-values is the extension of the accuracy function. Nevertheless, in the
literature several non-equivalent such “extensions” have been proposes. In [29], [30], it was made an
analysis of five of such proposals concluding that the more reasonable would be the new accuracy function
proposed in that paper and the one proposed in [66]. Here we will consider Xu’s accuracy function:
because, analogously to the case of 𝑆, the Xu’s accuracy function when applied to semi-diagonal elements
is the same, up to an isomorphisms 𝜓 and 𝜑, as ℎ∗
and ℎ, respectively, i.e. 𝐻(𝜓(𝑥)) = ℎ∗
(𝑥) for any 𝑥 ∈ 𝐿∗
and 𝐻(𝜑(𝑋)) = ℎ(𝑋) for any 𝑋 ∈ 𝕃. In addition, the range of 𝐻, ℎ and ℎ∗
are the same.
3.2| Order for 𝕃∗
-Values
In [56], it was introduced the notion of n-dimensional fuzzy interval and it was observed that 4-dimensional
fuzzy sets are isomorphic to IVAIFS. The degrees in an n-dimensional fuzzy interval take values in
𝐿𝑛([0,1]) = {(𝑥1, … , 𝑥𝑛) ∈ [0,1]𝑛
/𝑥𝑖 ≤ 𝑥𝑖+1 for each 𝑖 = 1,… , 𝑛 − 1}. In [9] the elements of 𝐿𝑛([0,1]) are called
n-dimensional intervals and the bijection 𝜚: 𝕃∗
→ 𝐿4([0,1]) defined by 𝜚(𝑋) = (𝛻(𝑋),△ (𝑋),1 −△ (𝑋̃ ),1 −
𝛻(𝑋̃ )) was provided. One of the possible interpretations considered in [9] for the 4-dimensional intervals
(𝑥1, 𝑥2, 𝑥3, 𝑥4) is that the intervals [𝑥1, 𝑥2] and [𝑥3, 𝑥4] represent an interval uncertainty in the bounds of an
interval-valued degree, i.e. of an element [𝑥, 𝑦] ∈ 𝕃, and so 𝑥 ∈ [𝑥1, 𝑥2] and 𝑦 ∈ [𝑥3, 𝑥4]. Having it in mind,
we introduce the notion of membership of 𝕃-values in 𝕃∗
-values.
Definition 5. Let X ∈ 𝕃 and 𝑋 ∈ 𝕃∗
. We say that 𝑋 ∈ 𝑋 if 𝑋 ∈ 𝑋 and 𝑋 ∈ 𝑋̃ 𝑐
.
Observe that this notion is strongly related to the notion of nesting given in
[6], [7] and therefore also can be used as a representation of IVAIFS by pairs of
AIFS.
Notice that, for each 𝑋, 𝑌, 𝑍 ∈ 𝕃,
I. if 𝑋 ⊆ 𝑌 ⊆ 𝑍 and 𝑋, 𝑍 ∈ 𝑌 for some 𝑌 ∈ 𝕃∗
, then 𝑌 ∈ 𝑌;
II. if 𝑋 ≤𝕃 𝑌 ≤𝕃 𝑍 and 𝑋, 𝑍 ∈ 𝑌 for some 𝑌 ∈ 𝕃∗
, then 𝑌 ∈ 𝑌;
III. 𝑌 ∈ 𝜑(𝑋) iff 𝑌 = 𝑋.
For any 𝑋 ∈ 𝕃∗
we will denote
i.e. 𝑋
⃗ = [𝛻(𝑋),1 −△ (𝑋̃ )] and 𝑋
⃖ = [△ (𝑋),1 − 𝛻(𝑋̃ )]. Notice that, the set 𝑆𝑋 = {𝑋 ∈ 𝕃/𝑋 ∈ 𝑋} is bounded, i.e.
for any 𝑋 ∈ 𝑆𝑋, 𝑋
⃗ ≤𝕃 𝑋 ≤𝕃 𝑋
⃖ and 𝑋
⃗, 𝑋
⃖ ∈ 𝑆𝑋. Thus, 𝑆𝑋 is a closed interval ([𝑋
⃗,𝑋
⃖]) of 𝕃-values and hence,
analogously to 𝕃-values, 𝕃∗
-values also have a dual nature: as an ordered pair of 𝕃-values with some
condition and as a set (an interval) of 𝕃-values.
3.2.1 | Subset order for 𝕃∗
-values
Since the usual membership relation is used to introduce the subset relation in set theory, the relation ∈
will allow us to introduce a notion of subset between 𝕃∗
-values. Let 𝑋, 𝑌 ∈ 𝕃∗
, we say that 𝑋 ⊆ 𝑌 if for
each 𝑋 ∈ 𝑋 we have that 𝑋 ∈ 𝑌. Analogously to the case of 𝕃-values, we can also define this inclusion
relation via the bounds of the interval associated to 𝕃∗
-values.
H(X) =
v(X) + v(X̃)
2
.
X
⃗ = [∇(X),∇(X̃ c
)] and X
⃖ = [△ (X),△ (X̃ c
)], (11)

248
Da
Silva
et
al.
|J.
Fuzzy.
Ext.
Appl.
2(3)
(2021)
239-261
Proposition 4. Let 𝑋, 𝑌 ∈ 𝕃∗
. Then the following expression are equivalents
I. 𝑋 ⊆ 𝑌;
II. 𝑆𝑋 ⊆ 𝑆𝑌;
III. 𝑌
⃗ ≤𝕃 𝑋
⃗ ≤𝕃 𝑋
⃖ ≤𝕃 𝑌
⃖;
IV. 𝑋 ⊆ 𝑌 𝑎𝑛𝑑 𝑋̃ ⊆ 𝑌̃.
Proof.
I. 1⇒2: If X▁(⊆) Y then for each X▁(∈) X also X▁(∈) Y, and so S_X⊆S_Y.
II. 2 ⇒ 3: Straightforward once that 𝑆𝑋 = [𝑋
⃗, 𝑋
⃖].
3 ⇒ 4 𝑌
⃗ ≤𝕃 𝑋
⃗ ≤𝕃 𝑋
⃖ ≤𝕃 𝑌
⃖ then by definition [𝛻(𝑌), 𝛻(𝑌̃𝑐
)] ≤𝕃 [𝛻(𝑋), 𝛻(𝑋̃ 𝑐
)] ≤𝕃 [△ (𝑋),△
(𝑋̃ 𝑐
)] ≤𝕃 [△ (𝑌),△ (𝑌̃𝑐
)] So, 𝛻(𝑌) ≤ 𝛻(𝑋) ≤△ (𝑋) ≤△ (𝑌) and 𝛻(𝑌̃𝑐
) ≤ 𝛻(𝑋̃ 𝑐
) ≤△ (𝑋̃ 𝑐
) ≤△ (𝑌̃𝑐
), i.e.
1 −△ (𝑌̃) ≤ 1 −△ (𝑋̃ ) ≤ 1 − 𝛻(𝑋̃ ) ≤ 1 − 𝛻(𝑌̃). Therefore 𝑋 ⊆ 𝑌 𝑎𝑛𝑑 𝑋̃ ⊆ 𝑌̃.
IV. 4 ⇒ 1: If 𝑋 ∈ 𝑋 then 𝑋 ∈ 𝑋 and 𝑋 ∈ 𝑋̃ 𝑐
. So, because 𝑋 ⊆ 𝑌 and 𝑋̃ ⊆ 𝑌̃, then 𝑋 ∈ 𝑌 and 𝑋 ∈ 𝑌̃𝑐
.
Therefore, 𝑋 ∈ 𝑌 and hence 𝑋 ⊆ 𝑌.
Remark 5. Some properties of ▁(⊆):
i. It is a partial order on 𝕃∗
-values;
ii. For each 𝑋, 𝑌 ∈ 𝕃, 𝜑(𝑋) ⊆ 𝜑(𝑌) iff 𝑋 = 𝑌;
iii. For each 𝑥, 𝑦 ∈ [0,1], 𝜙(𝑥) ⊆ 𝜙(𝑦) iff 𝑥 = 𝑦;
iv. Defining the complement of 𝕃∗
-values by 𝑋𝑐
= (𝑋̃ , 𝑋), then 𝑋 ⊆ 𝑌 iff 𝑋𝑐
⊆ 𝑌𝑐
.
3.2.2| Extension of ≤𝑿𝒀 total order for 𝕃∗
-values
In order to rank any possible set of 𝕃∗
-values it is necessary to provide a total order on 𝕃∗
, as made in
[68] for 𝐿∗
-values which was based on the score and accuracy index. Following the same idea, we define
the next binary relation on 𝕃∗
-values:
for any 𝑋, 𝑌 ∈ 𝕃∗
, where 𝑋 <𝑆 𝑌 iff 𝑋 ≤𝑆 𝑌 and 𝑋 ≡𝑆 𝑌.
Nevertheless, as noted in [64], this relation is not an order. However, in [64] it was provided the next
total order1 for 𝕃∗
:
for any 𝑋, 𝑌 ∈ 𝕃∗
, where 𝑇(𝑋) = 𝑤(𝑋) − 𝑤(𝑋̃ ) and 𝐺(𝑋) = 𝑤(𝑋) + 𝑤(𝑋̃ ).
1 In [64] was not claimed this, but from Proposition 4.1. of [31], it is possible to conclude that this order is total.
X ≤S,H Y iff {
X <S Y or
X ≡S Y and H(X) ≤ H(Y)
(12)
X ≼ Y iff
{
X <S Y or
X ≡S Y and H(X) < H(Y) or
X ≡S Y and H(X) = H(Y) and T(X) < T(Y) or
X ≡S Y and H(X) = H(Y) and T(X) = T(Y) and G(X) ≤ G(Y)
(13)

249
An
interval-valued
atanassov's
intuitionistic
fuzzy
multi-attribute
group
decision
making
method
based
on
the
best
representation
of
the
WA
and
OWA
operators
In [29], it was defined a new total order for 𝕃∗
-values, denoted here by ≾, which is based on the total order
for 𝐿∗
-values of Xu and Yager given by Eq. (3).
Theorem 1. [29] The binary relation ≾ on 𝕃∗
, defined for any 𝑋, 𝑌 ∈ 𝕃∗
by
is a total order.
Observe that the order ≾ is a particular instance of the admissible orders on 𝕃∗
introduced in [30], [31] (see
also [32]), i.e. is total and refines ≤𝕃∗.
Here, we propose a new total order, with the same principle as (14), but by considering other intervals:
Theorem 2. The binary relation ⪷ on 𝕃∗
, defined for any 𝑋, 𝑌 ∈ 𝕃∗
, by
is a total order.
Proof. Trivially, ⪷ is reflexive and antisymmetric. The transitivity of ⪷ follows from the transitivity of ≤𝑋𝑌
and equality. Analogously, the totallity of ⪷ follows from the totality of ≤𝑋𝑌.
4| 𝕃∗
-Representation of OWA
4.1 | 𝕃∗
-Representations of 𝕃-Functions
The notion of membership on 𝕃∗
-values also allows us to adapt the notion of interval representation for
𝕃∗
in the following way.
Definition 6. Let 𝐹: 𝕃𝑛
→ 𝕃 and ℱ : (𝕃∗
)𝑛
→ 𝕃∗
. ℱ is an 𝕃∗
-representation of 𝐹 if for each 𝑋𝑖 ∈ 𝕃∗
, and
𝑋𝑖 ∈ 𝑋𝑖, with 𝑖 = 1, … , 𝑛, F(𝑋1, … , 𝑋𝑛) ∈ ℱ(𝑋1, … , 𝑋𝑛).
Let 𝒢 , ℱ : (𝕃∗
)𝑛
→ 𝕃∗
. We say that ℱ is narrower than 𝒢 , denoted by 𝒢 ⊑𝕃∗ ℱ , if for any 𝑋𝑖 ∈ 𝕃∗
with
𝑖 = 1, … , 𝑛, ℱ (𝑋1, … , 𝑋𝑛) ⊆ 𝒢 (𝑋1, … , 𝑋𝑛). Analogously to the case of 𝕃-representation, we say that an 𝕃∗
-
representation ℱ of a function 𝐹: 𝕃𝑛
→ 𝕃 is better than another 𝕃∗
-representation 𝒢 of 𝐹 if 𝒢 ⊑𝕃∗ ℱ .
Theorem 3. Let 𝐹: 𝕃𝑛
→ 𝕃 be an isotone function. Then 𝐹̈: (𝕃∗
)𝑛
→ 𝕃∗
defined by
is an 𝕃∗
-representation of 𝐹. Moreover, if ℱ is another 𝕃∗
-representation of 𝐹 then ℱ ⊑𝕃∗ 𝐹̈.
Proof. Let 𝑋𝑖 ∈ 𝕃∗
with 𝑖 = 1, … , 𝑛. Since, 𝐹 is isotone w.r.t. ≤𝕃, then for each 𝑋𝑖 ∈ 𝑋𝑖 with 𝑖 = 1, … , 𝑛,
𝐹(𝑋1
⃗, … , 𝑋𝑛
⃗) ≤𝕃 𝐹(𝑋1, … , 𝑋𝑛) ≤𝕃 𝐹(𝑋1
⃖, … , 𝑋𝑛
⃖) and so 𝐹(𝑋1
⃗, … , 𝑋𝑛
⃗) ≤ 𝐹(𝑋1, … , 𝑋𝑛) ≤ 𝐹(𝑋1
⃖, … , 𝑋𝑛
⃖) and
𝐹(𝑋1
⃗, … , 𝑋𝑛
⃗) ≤ 𝐹(𝑋1, … , 𝑋𝑛) ≤ 𝐹(𝑋1
⃖, … , 𝑋𝑛
⃖). Therefore,
X ≾ Y iff X <XY Y or (X = Y and X̃ ≤XY Ỹ), (14)
X ⪷ Y iff X
⃗ <XY Y
⃗ or (X
⃗ = Y
⃗ and X
⃖ ≤XY Y
⃖). (15)
F̈(X1, … , Xn) = ([F(X1
⃗,… , Xn
⃗), F(X1
⃖, …, Xn
⃖)] , [1 − F(X1
⃖, … , Xn
⃖), 1 − F(X1
⃗, … , Xn
⃗)]), (16)

250
Da
Silva
et
al.
|J.
Fuzzy.
Ext.
Appl.
2(3)
(2021)
239-261
𝐹(𝑋1, … , 𝑋𝑛) ∈ [𝐹(𝑋1
⃗, … , 𝑋𝑛
⃗), 𝐹(𝑋1
⃖, … , 𝑋𝑛
⃖)] = 𝐹̈(𝑋1, … , 𝑋𝑛) and 𝐹(𝑋1, … , 𝑋𝑛) ∈
[𝐹(𝑋1
⃗, … , 𝑋𝑛
⃗), 𝐹(𝑋1
⃖, … , 𝑋𝑛
⃖)] = 𝐹̈(𝑋1, … , 𝑋𝑛)
̃ 𝑐
Hence, 𝐹(𝑋1, … , 𝑋𝑛) ∈ 𝐹̈(𝑋1, … , 𝑋𝑛).
If ℱ : (𝕃∗
)𝑛
→ 𝕃∗
is another 𝕃∗
-representation of 𝐹, then for each 𝑋𝑖 ∈ 𝕃∗
, and 𝑋𝑖 ∈ 𝑋𝑖, with 𝑖 = 1, … , 𝑛,
𝐹(𝑋1, … , 𝑋𝑛) ∈ ℱ (𝑋1, … , 𝑋𝑛). In particular, 𝐹(𝑋1
⃗, … , 𝑋𝑛
⃗), 𝐹(𝑋1
⃖, … , 𝑋𝑛
⃖) ∈ ℱ(𝑋1, … , 𝑋𝑛). So, by definition
of ∈, 𝐹(𝑋1
⃗, … , 𝑋𝑛
⃗), 𝐹(𝑋1
⃖, … , 𝑋𝑛
⃖) ∈ ℱ (𝑋1, … , 𝑋𝑛) and 𝐹(𝑋1
⃖, … , 𝑋𝑛
⃖), 𝐹(𝑋1
⃗, … , 𝑋𝑛
⃗) ∈ ℱ (𝑋1, … , 𝑋𝑛)
̃ 𝑐
, i.e. 1 −
𝐹(𝑋1
⃖, … , 𝑋𝑛
⃖), 1 − 𝐹(𝑋1
⃗ , … , 𝑋𝑛
⃗) ∈ ℱ (𝑋1, … , 𝑋𝑛)
̃ . Therefore, 𝐹̈(𝑋1, … , 𝑋𝑛) ⊆ ℱ (𝑋1, … , 𝑋𝑛) and
𝐹̈(𝑋1, … , 𝑋𝑛)
̃ ⊆ ℱ (𝑋1, … , 𝑋𝑛)
̃ and so, by Proposition 4, 𝐹̈(𝑋1, … , 𝑋𝑛) ⊆ ℱ (𝑋1, … , 𝑋𝑛). Hence, ℱ ⊑𝕃∗ 𝐹̈.
Corollary 1. Let F: 𝕃n
→ 𝕃 be an isotone function. Then
Proof. Straightforward from Theorem 3 and Eq. (11).
Corollary 2. Let f: [0,1]n
→ [0,1] be an isotone function. Then
Proof. Straightforward from Theorem 3 and eq. (11).
Corollary 3. Let 𝑓, 𝑔: [0,1]𝑛
→ [0,1] be isotone functions such that 𝑓 ≤ 𝑔. Then, 𝑓̂ ≤ 𝑔
̂, i.e.
𝑓̂(𝑋1, … , 𝑋𝑛) ≤𝕃∗ 𝑔
̂(𝑋1, … , 𝑋𝑛) for each 𝑋𝑖 ∈ 𝕃∗
with 𝑖 = 1, … , 𝑛.
Proof. Straightforward from Corollary 2 and definition of ≤𝕃∗.
Proposition 5. Let F: 𝕃n
→ 𝕃 be an isotone function. Then F̈(𝒟S) ⊆ 𝒟S and F̈(𝒟) ⊆ 𝒟
Proof. For any 𝑖 = 1, … , 𝑛, let 𝑋𝑖 ∈ 𝒟𝑆. Then 𝑋𝑖 = ([𝑥𝑖, 𝑥𝑖], [𝑦𝑖,𝑦𝑖]) for some 𝑥𝑖, 𝑦𝑖 ∈ [0,1] such that 𝑥𝑖 +
𝑦𝑖 ≤ 1. Since 𝑋𝑖
⃗ = [𝑥𝑖, 1 − 𝑦𝑖] = 𝑋𝑖
⃖ then, by Eq. (16), 𝐹̈(𝑋1, … , 𝑋𝑛) and 𝐹̈(𝑋1, … , 𝑋𝑛)
̃ are degenerate intervals
and so 𝐹̈(𝑋1, … , 𝑋𝑛) ∈ 𝒟𝑆.
For any 𝑖 = 1, … , 𝑛, let 𝑋𝑖 ∈ 𝒟. Then 𝑋𝑖 = ([𝑥𝑖,𝑥𝑖], [1 − 𝑥𝑖, 1 − 𝑥𝑖]) for some 𝑥𝑖 ∈ [0,1]. Since 𝑋𝑖
⃗ = [𝑥𝑖,𝑥𝑖] =
𝑋𝑖
⃖ then, by equation (16), 𝐹̈(𝑋1, … , 𝑋𝑛) and 𝐹̈(𝑋1, … , 𝑋𝑛)
̃ are degenerate intervals and 𝐹̈(𝑋1, … , 𝑋𝑛) =
𝐹̈(𝑋1, … , 𝑋𝑛)
̃ 𝑐
. So 𝐹̈(𝑋1, … , 𝑋𝑛) ∈ 𝒟 .
Lemma 1. Let X, Y ∈ 𝕃∗
. Then X
⃗ ⊆ Y
⃗ and X
⃖ ⊆ Y
⃖ iff X ≤ Y and X̃ ≤ Ỹ. Dually, X ⊆ Y and X̃ ⊆ Ỹ iff X
⃗ ≤
Y
⃗ and X
⃖ ≤ Y
⃖.
Proof. X
⃗ ⊆ Y
⃗ and X
⃖ ⊆ Y
⃖ iff ∇(Y) ≤ ∇(X), △ (Y) ≤△ (X), ∇(X̃ 𝑐
) ≤ ∇(Ỹ 𝑐
) and △ (X̃ 𝑐
) ≤△ (Ỹ 𝑐
) iff X ≤ Y
and X̃ 𝑐
≤ Ỹ 𝑐
iff X ≤ Y and X̃ ≤ Ỹ. The other case is analogous.
Proposition 6. Let F: 𝕃n
→ 𝕃 be an isotone function. Then
Proof. Straightforward from Lemma 1 and Corollary 1.
F̈(X1, … , Xn)
⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗ = F(X1
⃗, … , Xn
⃗) and F̈(X1, … , Xn)
⃖⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗ = F(X1
⃖ , … , Xn
⃖). (17)
f̂(X1, …, Xn) = (f̂(X1, …, Xn), f̂ (X1
̃
c
, …, Xn
̃
c
)c). (18)
F̈(X1, … , Xn) = F(X1, … , Xn) and F̈(X1, … , Xn)
̃ = F(X1
̃, … , Xn
̃). (19)

251
An
interval-valued
atanassov's
intuitionistic
fuzzy
multi-attribute
group
decision
making
method
based
on
the
best
representation
of
the
WA
and
OWA
operators
Proposition 7. Let F, G: 𝕃n
→ 𝕃 be isotone functions. If F ⊑𝕃 G then F̈ ⊑𝕃∗ G
̈ .
Proof. Let X𝑖 ∈ 𝕃∗
for any 𝑖 = 1, … , 𝑛. Since, 𝐹 ⊑𝕃 𝐺, then 𝐺(X1
⃗, … , X𝑛
⃗) ⊆ 𝐹(X1
⃗, … , X𝑛
⃗) and 𝐺(X1
⃖, … , X𝑛
⃖) ⊆
𝐹(X1
⃖, … , X𝑛
⃖) and so 𝐹(X1
⃗, … , X𝑛
⃗) ≤ 𝐺(X1
⃗, … , X𝑛
⃗) ≤ 𝐺(X1
⃗, … , X𝑛
⃗) ≤ 𝐹(X1
⃗, … , X𝑛
⃗) and 𝐹(X1
⃖, … , X𝑛
⃖) ≤
𝐺(X1
⃖, … , X𝑛
⃖) ≤ 𝐺(X1
⃖, … , X𝑛
⃖) ≤ 𝐹(X1
⃖, … , X𝑛
⃖). Therefore, by Theorem 3, 𝐹̈(X1, … , X𝑛) ⊆ 𝐺̈ (X1, … , X𝑛). Hence,
𝐹̈ ⊑𝕃∗ 𝐺̈ .
Note that, considering the interval point of view for 𝕃∗
-values, we have that
4.2 | 𝕃∗
-Representations of [𝟎, 𝟏]-Functions
Let 𝑥 ∈ [0,1] and X ∈ 𝕃∗
. Then 𝑥 ∈∗∗
X if 𝜙(𝑥) ⊆ X, i.e. if 1 − ∇(X̃) ≤ 𝑥 ≤△ (X). There is a close relation
between ∈ and ∈∗∗
as can we see in the next proposition.
Proposition 8. Let 𝑋 ∈ 𝕃∗
and 𝑋 ∈ 𝕃. 𝑋 ∈ 𝑋 if and only if 𝑋 ∈∗∗
𝑋 and 𝑋 ∈∗∗
𝑋
Proof. Since, trivially, 𝜙(𝑥)
⃗ = [𝑥, 𝑥] = 𝜙(𝑥)
⃖ for any 𝑥 ∈ 𝕃, then
𝑋 ∈ X 𝑖𝑓𝑓 X
⃗ ≤𝕃 𝑋 ≤𝕃 X
⃖
𝑖𝑓𝑓 X
⃗ ≤𝕃 [𝑋,𝑋] ≤𝕃 [𝑋, 𝑋] ≤𝕃 X
⃖
𝑖𝑓𝑓 𝜙(𝑋) ⊆ X 𝑎𝑛𝑑 𝜙(𝑋) ⊆ X 𝑏𝑦 𝑃𝑟𝑜𝑝. 4
𝑖𝑓𝑓 𝑋 ∈∗∗
X 𝑎𝑛𝑑 𝑋 ∈∗∗
X 𝑏𝑦 𝑑𝑒𝑓. 𝑜𝑓 ∈∗∗
With this notion of membership, we can naturally extend the notion of 𝕃-representation of fuzzy function
for the 𝕃∗
-representation of fuzzy function and introduce a new notion of inclusion for 𝕃∗
-values.
Definition 7. Let f: [0,1]n
→ [0,1] and ℱ : (𝕃∗
)n
→ 𝕃∗
. ℱ is an 𝕃∗
-representation of f if for each Xi ∈ 𝕃∗
and xi ∈∗∗
Xi, with i = 1, … , n, we have that f(x1, … , xn) ∈∗∗
ℱ (X1, … , Xn)
Let X, Y ∈ 𝕃∗
. Then X ⊆∗∗
Y if for each 𝑥 ∈∗∗
X, also 𝑥 ∈∗∗
Y. However, ⊆∗∗
is not a partial order (it is not
antisymmetric – e.g. consider X = ([0.2,0.3], [0.4,0.5]) and Y = ([0.1,0.3],[0.2,0.5])). Therefore, we just
consider ⊆ as the extension of inclusion order for 𝕃∗
.
Analogously to the case of 𝕃-representation, we said that an 𝕃∗
-representation ℱ of a function 𝑓: [0,1]𝑛
→
[0,1] is better than another 𝕃∗
-representation 𝒢 of 𝑓 if 𝒢 ⊑𝕃∗ ℱ .
Proposition 9. Let f: [0,1]n
→ [0,1] and F: 𝕃n
→ 𝕃 be isotone functions. If F is an 𝕃-representation of f
then F̈ is an 𝕃∗
-representation of f.
Proof. If 𝑥𝑖 ∈∗∗
X𝑖 for any 𝑖 = 1,… , 𝑛, then 𝜙(𝑥𝑖) = ([𝑥𝑖,𝑥𝑖], [1 − 𝑥, 1 − 𝑥]) ⊆ X𝑖 and so, by Proposition 4,
[𝑥𝑖,𝑥𝑖] ⊆ X𝑖 and [𝑥𝑖, 𝑥𝑖]𝑐
⊆ X𝑖
̃, or equivalently, [𝑥𝑖,𝑥𝑖] ⊆ X𝑖
̃
𝑐
. Therefore, 𝑥𝑖 ∈ X𝑖
̃ and 𝑥𝑖 ∈ X𝑖
̃
𝑐
. Thus, since 𝐹 is
an 𝕃-representation of 𝑓, 𝑓(𝑥1, … , 𝑥𝑛) ∈ 𝐹(X1, … , X𝑛) and so [𝑓(𝑥1, … , 𝑥𝑛), 𝑓(𝑥1, … , 𝑥𝑛)] ⊆ 𝐹(X1, … , X𝑛) and
[𝑓(𝑥1, … , 𝑥𝑛), 𝑓(𝑥1, … , 𝑥𝑛)]𝑐
⊆ 𝐹(X1
̃
𝑐
, … , X𝑛
̃
𝑐
)𝑐
. Hence, by Corollary 2, [𝑓(𝑥1, … , 𝑥𝑛), 𝑓(𝑥1, … , 𝑥𝑛)] ⊆
𝐹̈(X1, … , X𝑛) and [𝑓(𝑥1, … , 𝑥𝑛), 𝑓(𝑥1, … , 𝑥𝑛)]𝑐
⊆ 𝐹̈(X1, … , X𝑛)
̃ . Therefore, 𝜙(𝑓(𝑥1, … , 𝑥𝑛)) ⊆ 𝐹̈(X1, … , X𝑛), i.e.
𝑓(𝑥1, … , 𝑥𝑛) ∈∗∗
𝐹̈(X1, … , X𝑛). So, 𝐹̈ is an 𝕃∗
-representation of 𝑓.
F̈(X1, … , Xn) ≃ [F(X1
⃗, … , Xn
⃗), F(X1
⃖, … , Xn
⃖)].

252
Da
Silva
et
al.
|J.
Fuzzy.
Ext.
Appl.
2(3)
(2021)
239-261
Theorem 4. Let f: [0,1]n
→ [0,1] be an isotone function. f̂ is the best, w.r.t. ⊑𝕃∗, 𝕃∗
-representation of f.
Proof. From Propositions 1 and 9 and Remark 2 it follows that 𝑓̂ is an 𝕃∗
-representation of 𝑓. Thus, it only
remains to prove that is the best one.
Let ℱ : (𝕃∗
)𝑛
→ 𝕃∗
be another 𝕃∗
-representation of 𝑓 and X𝑖 ∈ 𝕃∗
for 𝑖 = 1, … , 𝑛. If 𝑋𝑖 ∈ X𝑖, for any 𝑖 =
1,… , 𝑛, then by Proposition 8 𝑋𝑖 ∈∗∗
X𝑖 and 𝑋𝑖 ∈∗∗
X𝑖. So, because ℱ is 𝕃∗
-representation of 𝑓,
𝑓(𝑋1, … , 𝑋𝑛) ∈∗∗
ℱ (X1, … , X𝑛) and 𝑓(𝑋1, … , 𝑋𝑛) ∈∗∗
ℱ (X1, … , X𝑛). Thus, by equation (5),
𝑓̂(𝑋1, … , 𝑋𝑛) ∈∗∗
ℱ (X1, … , X𝑛) and 𝑓̂(𝑋1, … , 𝑋𝑛) ∈∗∗
ℱ (X1, … , X𝑛). Therefore, by Proposition 8,
𝑓̂(𝑋1, … , 𝑋𝑛) ∈ ℱ (X1, … , X𝑛), i.e. ℱ is an 𝕃∗
-representation of 𝑓̂. Hence, by Theorem 3, 𝐹 ⊑𝕃∗ 𝑓̂, and so
𝑓
̂ is a better 𝕃∗
-representation of 𝑓 than ℱ .
4.3| The Best 𝕃∗
-Representation of the OWA Operator
Aggregation functions play an important role in fuzzy sets theory, so it is natural to extend this definition
for IVAIFS.
Definition 8. An n-ary function 𝒜 : (𝕃∗
)n
→ 𝕃∗
is an n-ary interval-valued Atanassov’s intuitionistic
aggregation function if
I. If X𝑖 ≤𝕃∗ Y𝑖, for each 𝑖 = 1, … , 𝑛, then 𝒜(X1, … , X𝑛) ≤𝕃∗ 𝒜(Y1, … , Y𝑛);
II. 𝒜 (0𝕃∗, … , 0𝕃∗) = 0𝕃∗ and 𝒜 (1𝕃∗,… , 1𝕃∗) = 1𝕃∗.
Theorem 5. Let 𝐴: [0,1]𝑛
→ [0,1] be an n-ary aggregation function. Then 𝐴
̂ is an n-ary interval-valued
Atanassov’s intuitionistic aggregation function. Moreover, if A is idempotent and/or symmetric, then 𝐴
̂
is also idempotent and/or symmetric.
Proof. Straightforward from Corollary 2 and Remark 2.
In order to motivate the next section, we will need some arithmetic operations on 𝕃∗
.
Scalar product. The multiplication ⊙ of an scalar 𝜆 ∈ [0,1] by X ∈ 𝕃∗
is defined by
Division by a positive integer. Let 𝑛 ∈ ℤ+
be a positive integer, then
X
𝑛
=
1
𝑛
⊙ X
Limited addition. Let 𝑋, 𝑌 ∈ 𝐿∗
. Then
It is clear that these operations are well defined, i.e. they always provide an element of 𝕃∗
.
Definition 9. Let 𝛬 be an n-ary weighting vector, i.e. 𝛬 = (𝜆1, … , 𝜆𝑛) ∈ [0,1]𝑛
such that ∑𝑛
𝑖=1 𝜆𝑖 = 1. The
n-dimensional interval-valued intuitionistic weighted average 𝕃∗
− 𝑊𝐴𝛬 is given by
where the sum is w.r.t. the limited addition.
λ ⊙ X = (λX, λX̃). (20)
X ⊕ Y = (X[+]Y,X̃[+]Ỹ). (21)
𝕃∗
− WA Λ(X1, … , Xn) = ∑n
i=1 λi ⊙ Xi, (22)

253
An
interval-valued
atanassov's
intuitionistic
fuzzy
multi-attribute
group
decision
making
method
based
on
the
best
representation
of
the
WA
and
OWA
operators
Lemma 2. Let 𝑋, 𝑌 ∈ 𝕃∗
and 𝜆1, 𝜆2 ∈ [0,1] such that 𝜆1 + 𝜆2 ≤ 1. Then 𝜆1 ⊙ 𝑋 ⊕ 𝜆2 ⊙ 𝑌 = (𝜆1𝑋 +
𝜆2𝑌, 𝜆1𝑋̃ + 𝜆2𝑌̃).
Proof. Straightforward from Eqs. (7), (20) and (21).
Lemma 3. Let 𝛬 be a weighting vector. Then, waΛ
̂ (X1
c
,… , Xn
c
)c
= waΛ
̂ (X1, … , Xn).
Proof. Straightforward from Proposition 1 and the fact that 1 − 𝑤𝑎Λ(1 − 𝑥1, … ,1 − 𝑥𝑛) = 𝑤𝑎Λ(𝑥1, … , 𝑥𝑛).
Theorem 6. Let 𝛬 be a weighting vector. Then 𝕃∗
− 𝑊𝐴𝛬 = 𝑤𝑎𝛬
̂ , i.e. is the best 𝕃∗
-representation of the
weighted average operator.
Proof. First note that by the monotonicity of the weighted average operator, 𝑤𝑎Λ
̂ (𝑋1, … , 𝑋𝑛) =
[𝑤𝑎Λ(𝑋1,… , 𝑋𝑛), 𝑤𝑎Λ(𝑋1, … , 𝑋𝑛)] = ∑𝑛
𝑖=1 𝜆𝑖𝑋𝑖. So,
𝑤𝑎Λ
̂ (X1, … , X𝑛) = (𝑤𝑎Λ
̂ (X1, … , X𝑛), 𝑤𝑎Λ
̂ (X1
𝑐
̃, … , X𝑛
𝑐
̃)𝑐
) 𝑏𝑦 𝐶𝑜𝑟. 2
= (𝑤𝑎Λ
̂ (X1, … , X𝑛), 𝑤𝑎Λ
̂ (X1
̃, … , X𝑛
̃)) 𝑏𝑦 𝐿𝑒𝑚𝑚𝑎 23
= (∑𝑛
𝑖=1 𝜆𝑖X𝑖,∑𝑛
𝑖=1 𝜆𝑖X𝑖
̃) 𝑏𝑦 𝑃𝑟𝑜𝑝. 231
= ∑𝑛
𝑖=1 𝜆𝑖 ⊙ X𝑖 𝑏𝑦 𝐿𝑒𝑚𝑚𝑎 2312 .
Definition 10. Let 𝛬 be an n-ary weighting vector, i.e. 𝛬 = (𝜆1, … , 𝜆𝑛) ∈ [0,1]𝑛
such that ∑𝑛
𝑖=1 𝜆𝑖 = 1. The
n-dimensional interval-valued intuitionistic weighted addition 𝕃∗
− 𝑂𝑊𝐴𝛬 is given by
where the sum is w.r.t. the limited addition and
with 𝛾𝑗: {0,1 … , 𝑛} → {0,1 … , 𝑛} for 𝑗 = 1, … ,4, being permutations such that 𝛻(𝑋𝛾1(𝑖)) ≥ 𝛻(𝑋𝛾1(𝑖+1)), △
(X𝛾2
(𝑖)) ≥△ (X𝛾2
(𝑖+1)), ∇(X𝛾3(𝑖)
̃) ≤ ∇(X𝛾3(𝑖+1)
̃) and △ (X𝛾4(𝑖)
̃) ≤△ (X𝛾4(𝑖+1)
̃) for any 𝑖 = 1, … , 𝑛 − 1.
Lemma 4. Let 𝛬 be a weighting vector. Then, 𝑜𝑤𝑎𝛬
̂(𝑋1
𝑐
, … , 𝑋𝑛
𝑐
)𝑐
= 𝑜𝑤𝑎𝛬𝑟
̂ (𝑋1, … , 𝑋𝑛) where 𝛬𝑟
=
(𝜆𝑛,… , 𝜆1).
Proof. Straightforward from Proppsition 1 and the fact that 1 − 𝑜𝑤𝑎Λ(1 − 𝑥1, … ,1 − 𝑥𝑛) = 𝑜𝑤𝑎Λ𝑟(𝑥1, … , 𝑥𝑛).
Theorem 7. Let 𝛬 be a weighting vector. Then 𝕃∗
− 𝑂𝑊𝐴𝛬 = 𝑜𝑤𝑎𝛬
̂, i.e. is the best 𝕃∗
-representation of
the ordered weighted average operator.
Proof.
𝕃∗
− OWA Λ(X1, … , Xn) = ∑
n
i=1
λi ⊙ Xγ(i), (23)
X𝛾(𝑖) = ([∇(X𝛾1(𝑖)),△ (X𝛾2(𝑖))],[∇(X𝛾3(𝑖)
̃),△ (X𝛾4(𝑖)
̃)]). (24)

254
Da
Silva
et
al.
|J.
Fuzzy.
Ext.
Appl.
2(3)
(2021)
239-261
Corollary 4. 𝕃∗
− 𝑂𝑊𝐴𝛬 is an idempotent and symmetric n-ary interval-valued Atanassov’s
intuitionistic aggregation function. In addition, 𝕃∗
− 𝑂𝑊𝐴𝛬 is bounded, i.e. 𝑚𝑖𝑛
̂ ≤𝕃∗ 𝕃∗
−
𝑂𝑊𝐴Λ ≤𝕃∗ 𝑚𝑎𝑥
̂
Proof. Straightforward from Theorems 7 and 5 and Corollary 3.
5| A Method for Multi-attribute Group Decision Making Based
Interval-Valued Atanassov’s Intuitionistic Decision Matrices
Let 𝐸 = {𝑒1, … , 𝑒𝑚} be a set of experts, 𝑋 = {𝑥1, … , 𝑥𝑛} be a finite set of alternatives, and 𝐴 = {𝑎1, … , 𝑎𝑝}
be a set of attributes or criteria. The decision makers determines a weighting vector 𝑊 = (𝑤1, … , 𝑤𝑝)𝑇
for the attributes. A method for MAGDM based on IVAIDM is an algorithm which determines a
ranking of the alternatives in 𝑋 based in the opinion of each expert in 𝐸 of how much the alternatives
attend each attribute. In particular we consider the case where the evaluation of the experts contains
imprecision and hesitation which is represented by interval-valued Atanassov’s intuitionistic degrees.
We propose the next method (algorithm) to obtain such ranking:
𝑋, 𝑊, and for every 𝑙 = 1,… , 𝑚 an 𝕃∗
-valued decision matrix 𝑅𝑙
of dimension 𝑛 × 𝑝 where each position
(𝑖,𝑗) in 𝑅𝑙
, denoted by 𝑅𝑖𝑗
𝑙
, contains the interval-valued Atanassov’s intuitionistic value which reflects how
much the alternative 𝑥𝑖 attends the attribute (or criterium1) 𝑎𝑗.
A ranking 𝑟: 𝑋 → {1, … , 𝑛}, denoting that an alternative 𝑥 ∈ 𝑋 is better than an alternative 𝑦 ∈ 𝑋 whenever
𝑟(𝑥) ≤ 𝑟(𝑦) and when 𝑟(𝑥) = 𝑟(𝑦) meaning that the method is not able of determine if 𝑥 is better or worst
alternative than 𝑦2.
1 For the case of the cost criteria is considered the usal complement of these interval-valued Atanassov’s intuitionistic values.
2 The most decision making methods admits cases for which the method is unable of discriminate between two different alternatives
which is better.
𝕃∗
− OWA Λ(X1, … , Xn)
= ∑
n
i=1
λi ⊙ Xγ(i) by eq. (23)
= 𝕃∗
− WA Λ(Xγ(1),… , Xγ(n)) by eq. (2322)
= waΛ
̂ (Xγ(1), … , Xγ(n)) by Thm. 23226
= (waΛ
̂ (Xγ(1),… , Xγ(n)),waΛ
̂ (Xγ(1)
̃
c
, … , Xγ(n)
̃
c
)c
) by eq. (2322618)
= (waΛ
̂ (Xγ(1),… , Xγ(n)),waΛ
̂ (Xγ(1)
̃,… , Xγ(n)
̃)) by Lemma 23226183
= (waΛ
̂ ([∇(Xγ1(1)),△ (Xγ2(1))],… , [∇(Xγ1(n)),△ (Xγ2(n))],
waΛ
̂ ([∇(Xγ3(1)
̃ ),△ (Xγ4(1)
̃ )],… , [∇(Xγ3(n)
̃ ),△ (Xγ4(n)
̃ )])) by eq. (2322618324)
= ([waΛ(∇(Xγ1(1)),… , ∇(Xγ1(n))),waΛ(△ (Xγ2(1)),… ,△ (Xγ2(n)))],
[waΛ(∇(Xγ3(1)
̃ ),… , ∇(Xγ3(n)
̃ )),waΛ(△ (Xγ4(1)
̃ ),… ,△ (Xγ4(n)
̃ ))]) by eq. (23226183245)
= ([waΛ(∇(Xγ1(1)),… , ∇(Xγ1(n))),waΛ(△ (Xγ2(1)),… ,△ (Xγ2(n)))],
[waΛr(∇(Xγ3(n)
̃ ),… , ∇(Xγ3(1)
̃ )),waΛr(△ (Xγ4(n)
̃ ),… ,△ (Xγ4(1)
̃ ))])
= ([owaΛ(∇(Xγ(1)),… , ∇(Xγ(n))), owaΛ(△ (Xγ(1)), … ,△ (Xγ(n)))],
[owaΛr(∇(Xγ(1)
̃),… , ∇(Xγ(n)
̃)), owaΛr(△ (Xγ(1)
̃), … ,△ (Xγ(n)
̃))]) by eq. (232261832456)
= (owaΛ
̂ (Xγ(1),… , Xγ(n)),owaΛr
̂ (Xγ(1)
̃,… , Xγ(n)
̃)) by eq. (2322618324565)
= (owaΛ
̂ (X1, … , Xn), owaΛ
̂ (X1
̃
c
,… , Xn
̃
c
)c
) by Lemma 23226183245654
= owaΛ
̂ (X1, … , Xn). by eq. (2322618324565418)

255
An
interval-valued
atanassov's
intuitionistic
fuzzy
multi-attribute
group
decision
making
method
based
on
the
best
representation
of
the
WA
and
OWA
operators
Step 1. Aggregate the IVAIDM of all experts in a single IVAIDM ℛ𝒞, for each 𝑖 = 1, … , 𝑛 and 𝑗 = 1,… , 𝑝,
as follows:
where 𝛬 = (𝜆1, … , 𝜆𝑚) is the following weighting vector:
1. Case 𝑚 is even: 𝜆𝑖 =
1
2
𝑚
2
+2−𝑖
+
1
𝑚2
𝑚
2
for each 𝑖 = 1,… ,
𝑚
2
, and 𝜆𝑖 = 𝜆𝑚+1−𝑖 for each 𝑖 =
𝑚
2
+ 1, … , 𝑚.
2. Case 𝑚 is odd: 𝜆𝑖 =
1
2
𝑚+1
2
+2−𝑖
+
1
𝑚2
𝑚+1
2
+
1
4𝑚
for each 𝑖 = 1, … ,
𝑚+1
2
, and 𝜆𝑖 = 𝜆𝑚+1−𝑖 for each 𝑖 =
𝑚+1
2
+ 1, … , 𝑚.
Table 1. Assesses of expert 𝐩𝟏.
ℛ𝒞 is the IVAIDM of consensus of all expert opinions1.
Step 2. For each alternative 𝑥𝑖, with 𝑖 = 1,… , 𝑛, using 𝑤𝑎𝑊
̂ , determine the collective overall index 𝕃∗
-value
𝑂𝑖 as follows:
Step 3. Rank the alternatives by considering a total order on their collective overall index 𝕃∗
-values and
choosing the greatest one. Thus, the output function 𝑟: 𝑋 → {1, … , 𝑛} is defined by 𝑟(𝑥𝑖) = 𝑗 iff 𝑂𝑖 is the 𝑗th
greatest collective overall index. Notice that if two or more alternatives, e.g. 𝑥 and 𝑦, have the same
collective overall index, then 𝑟(𝑥) = 𝑟(𝑦).
Example 1. Consider the air-condition system selection problem used as example in [62]. This problem
considers three air-condition systems (alternatives) {𝐴1,𝐴2, 𝐴3}; four attributes: 𝑎1 (economical), 𝑎2
(function), 𝑎3 (being operative) and 𝑎4 (longevity); and three experts {𝑝1, 𝑝2, 𝑝3}. By using statistical methods,
for each expert 𝑝𝑙, alternative 𝐴𝑖 and atribute 𝑎𝑗 an interval-valued membership degree and an interval-
valued non-membership degree, i.e. an IVAIFV, is provided. These IVAIFV are summarized in the Tables
1, 2 and 3 (the same used in [62]). We consider the weighting vector 𝑊 = (0.2134,0.1707,0.2805,0.3354) for
the attributes2
.
Since we have three experts (𝑚 = 3), then the weighting vector 𝛬 is calculated as following:
𝜆1 =
1
23
+
1
3 ⋅ 22
+
1
4 ⋅ 3
=
1
8
+
1
6
= 0.2916,
𝜆2 =
1
22
+
1
3 ⋅ 22
+
1
4 ⋅ 3
=
1
4
+
1
6
= 0.416,
𝜆3 =
1
23
+
1
3 ⋅ 22
+
1
4 ⋅ 3
=
1
8
+
1
6
= 0.2916.
1 It is not hard of prove that when 𝑛 > 1, Λ is a weighting vector.
2 In [62] was considered the weights V = (0.35,0.28,0.46,0.55) which not satisfy the condition that the sums of the weights must be equal
to 1. W is the weighting vector obtained normalizing 𝑉 in order to satisfy this condition.
ℛ 𝒞ij = owaΛ
̂ (Rij
1
, … , Rij
m
). (25)
𝐑𝟏 𝐚𝟏 𝐚𝟐 𝐚𝟑 𝐚𝟒
A 1 ([0.4,0.8],[0.0,0.1]) ([0.3,0.6],[0.0,0.2]) ([0.2,0.7],[0.2,0.3]) ([0.3,0.4],[0.4,0.5])
A 2 ([0.5,0.7],[0.1,0.2]) ([0.3,0.5],[0.2,0.4]) ([0.4,0.7],[0.0,0.2]) ([0.1,0.2],[0.7,0.8])
A 3 ([0.5,0.7],[0.2,0.3]) ([0.6,0.8],[0.1,0.2]) ([0.4,0.7],[0.1,0.2]) ([0.6,0.8],[0.0,0.2])
Oi = waW
̂(ℛ 𝒞i1,… , ℛ 𝒞in). (26)

256
Da
Silva
et
al.
|J.
Fuzzy.
Ext.
Appl.
2(3)
(2021)
239-261
Table 2. Assesses of expert 𝐩𝟐.
Table 3. Assesses of expert 𝐩𝟑.
The Table 4 present the collective reflexive IvIFPR obtained from Tables 1, 2 and 3 by consider the Eq.
(25).
The collective overall preference obtained by using the calculation in Eq. (26), is the following:
𝑂1 = ([0.3509555488,0.6721],[0.140916,0.2651]),
𝑂2 = ([0.3867014634,0.6262],[0.180441,0.3184]),
𝑂3 = ([0.4086795732,0.6848],[0.111192,0.2443]).
Thus, considering this collective overall preference and the total orders shows in section III.B, we have
the ranking of the alternatives in the Table 5. Therefore, all the ranking obtained with this method, for
the different the orders considered, agree with four of the five ranking obtained in [39], [62], [63], for
this same illustrative example.
Example 2. Consider the investment choice problem used as example in [59], [60]. This problem
considers an investment company which would like to invest a sum of money in the best option among
the following five possible alternatives to invest the money: 𝐴1 is a car company; 𝐴2 is a food company;
𝐴3 is a computer company; 𝐴4 is an arms company; and 𝐴5 is a TV company. The choice of the best
investmente must be made taking into account the following four benefit criteria: c1 is the profit ability;
𝑐2 is the growth analysis; 𝑐3 is the social-political impact; and 𝑐4 is the enterprise culture. The five possible
alternatives will be evaluated considering the interval-valued intuitionistic fuzzy information given by
three decision makers 𝑒1, 𝑒2 and 𝑒3, who evaluate how much the alternative satisfies each one of the
criterias. These informations are summarized in the Tables 6, 7 and 8 (the same considered in [58], [59],
[60]).
Since, in [59], [60] it was not considered a weight for the criteria, here we consider that all criteria have
the same weight, i.e. we consider 𝑊 = (0.25,0.25,0.25,0.25). The ranking obtained by using our method
considering the four total orders and the obtained by [59], [60] is summarized in the Table 9.
Table 4. Collective reflexive IvIFPR.
𝐑𝟐 𝐚𝟏 𝐚𝟐 𝐚𝟑 𝐚𝟒
A1 ([0.5,0.9],[0.0,0.1]) ([0.4,0.5],[0.3,0.5]) ([0.5,0.8],[0.0,0.1]) ([0.4,0.7],[0.1,0.2])
A2 ([0.7,0.8],[0.1,0.2]) ([0.5,0.6],[0.2,0.3]) ([0.5,0.8],[0.0,0.2]) ([0.5,0.6],[0.3,0.4])
A3 ([0.5,0.6],[0.1,0.4]) ([0.6,0.7],[0.1,0.2]) ([0.4,0.8],[0.1,0.2]) ([0.2,0.6],[0.2,0.3])
𝐑𝟑 𝐚𝟏 𝐚𝟐 𝐚𝟑 𝐚𝟒
A1 ([0.3,0.9],[0.0,0.1]) ([0.2,0.5],[0.1,0.4]) ([0.4,0.7],[0.1,0.2]) ([0.3,0.6],[0.3,0.4])
A2 ([0.3,0.8],[0.1,0.2]) ([0.5,0.6],[0.1,0.3]) ([0.2,0.8],[0.0,0.2]) ([0.3,0.5],[0.2,0.3])
A3 ([0.2,0.6],[0.1,0.2]) ([0.2,0.6],[0.2,0.3]) ([0.3,0.6],[0.1,0.3]) ([0.4,0.7],[0.1,0.2])
𝐑𝐂 𝐚𝟏 𝐚𝟐 𝐚𝟑 𝐚𝟒
A 1 ([0.4,0.871],[0.0,0.1]) ([0.3,0.53],[0.13,0.371]) ([0.371,0.73],[0.1,0.2]) ([0.33,0.571],[0.271,0.371])
A 2 ([0.5,0.771],[0.1,0.2]) [0.4416,0.571],[0.171,0.33]) [0.371,0.771],[0.0,0.2]) [0.3,0.4416],[0.3875,0.4874])
A 3 [0.4125,0.63],[0.13,0.3]) ([0.4833,0.7],[0.13,0.23]) ([0.371,0.7],[0.1,0.23]) ([0.4,0.7],[0.1,0.23])

257
An
interval-valued
atanassov's
intuitionistic
fuzzy
multi-attribute
group
decision
making
method
based
on
the
best
representation
of
the
WA
and
OWA
operators
Table 5. Ranking obtained for the alternatives considering several total orders and the obtained in [62],
[63].
Table 6. Assesses of expert 𝐞𝟏.
Table 7. Assesses of expert 𝐞𝟐.
Table 8. Assesses of expert 𝐞𝟑.
Table 9. Ranking obtained for the alternatives considering several total orders and the obtained in [62].
Thus, making an analysis of these rankings of the alternatives we have that there is an absolute consensus
that the worst alternative is A1 and the second worst alternative is A4. On the other hand, if we consider,
for the other alternatives, the amount of times that an alternative was better ranked than the others (which
is summarized in the Table 10) we can conclude that the more rasonable ranking of the alternatives would
be𝐴5 > 𝐴2 > 𝐴3 > 𝐴4 > 𝐴1
which agrees with the ranking obtained in [60] and also for the proposed method with the orders ⪅ and ≾.
This way of aggregate or fuses many rankings of a set of alternatives corresponds to the ranking fusion
function M2 of [20].
⪅ ≾ ≼ [62] [63] (a), (b) and (d) [63] (c)
A 3 A 3 A 3 A 3 A 3 A 2
A 1 A 1 A 1 A 1 A 1 A 1
A 2 A 2 A 2 A 2 A 2 A 3
𝐑𝟏 𝐚𝟏 𝐚𝟐 𝐚𝟑 𝐚𝟒
A1 ([0.4,0.5],[0.3,0.4]) ([0.4,0.6],[0.2,0.4]) ([0.1,0.3],[0.5,0.6]) ([0.3,0.4],[0.3,0.5])
A2 ([0.6,0.7],[0.2,0.3]) ([0.6,0.7],[0.2,0.3]) ([0.4,0.7],[0.1,0.2]) ([0.5,0.6],[0.1,0.3])
A3 ([0.6,0.7],[0.1,0.2]) ([0.5,0.6],[0.3,0.4]) ([0.5,0.6],[0.1,0.3]) ([0.4,0.5],[0.2,0.4])
A4 ([0.3,0.4],[0.2,0.3]) ([0.6,0.7],[0.1,0.3]) ([0.3,0.4],[0.1,0.2]) ([0.3,0.7],[0.1,0.2])
A5 ([0.7,0.8],[0.1,0.2]) ([0.3,0.5],[0.1,0.3]) ([0.5,0.6],[0.2,0.3]) ([0.3,0.4],[0.5,0.6])
𝐑𝟐 𝐚𝟏 𝐚𝟐 𝐚𝟑 𝐚𝟒
A 1 ([0.3,0.4],[0.4,0.5]) ([0.5,0.6],[0.1,0.3]) ([0.4,0.5],[0.3,0.4]) ([0.4,0.6],[0.2,0.4])
A 2 ([0.3,0.6],[0.3,0.4]) ([0.4,0.7],[0.1,0.2]) ([0.5,0.6],[0.2,0.3]) ([0.6,0.7],[0.2,0.3])
A 3 ([0.6,0.8],[0.1,0.2]) ([0.5,0.6],[0.1,0.2]) ([0.5,0.7],[0.2,0.3]) ([0.1,0.3],[0.5,0.6])
A 4 ([0.4,0.5],[0.3,0.5]) ([0.5,0.8],[0.1,0.2]) ([0.2,0.5],[0.3,0.4]) ([0.4,0.7],[0.1,0.2])
A 5 ([0.6,0.7],[0.2,0.3]) ([0.6,0.7],[0.1,0.2]) ([0.5,0.7],[0.2,0.3]) ([0.6,0.7],[0.1,0.3])
𝐑𝟑 𝐚𝟏 𝐚𝟐 𝐚𝟑 𝐚𝟒
A 1 ([0.2,0.5],[0.3,0.4]) ([0.4,0.5],[0.1,0.2]) ([0.3,0.6],[0.2,0.3]) ([0.3,0.7],[0.1,0.3])
A 2 ([0.2,0.7],[0.2,0.3]) ([0.3,0.6],[0.2,0.4]) ([0.4,0.7],[0.1,0.2]) ([0.5,0.8],[0.1,0.2])
A 3 ([0.5,0.6],[0.3,0.4]) ([0.7,0.8],[0.1,0.2]) ([0.5,0.6],[0.2,0.3]) ([0.4,0.5],[0.3,0.4])
A 4 ([0.3,0.6],[0.2,0.4]) ([0.4,0.6],[0.2,0.3]) ([0.1,0.4],[0.3,0.6]) ([0.3,0.7],[0.1,0.2])
A 5 ([0.6,0.7],[0.1,0.3]) ([0.5,0.6],[0.3,0.4]) ([0.5,0.6],[0.2,0.3]) ([0.5,0.6],[0.2,0.4])
Proposed Method The methods proposed in [59], [60]
⪅ ≾ ≼ [60] [59] 𝛾 < 0.378 [59] 𝛾 = 0.378 [59] 0.378 < 𝛾 <
1
[59] 𝛾 = 1
A 5 A 5 A 2 A 5 A 3 A 3 ∼ A 5 A 5 A 5
A 2 A 2 A 5 A 2 A 5 A 3 A 3 ∼ A 2
A 3 A 3 A 3 A 3 A 2 A 2 A 2
A 4 A 4 A 4 A 4 A 4 A 4 A 4 A 4
A 1 A 1 A 1 A 1 A 1 A 1 A 1 A 1

258
Da
Silva
et
al.
|J.
Fuzzy.
Ext.
Appl.
2(3)
(2021)
239-261
6| Final remarks
This paper proposes a new extension of the OWA and WA operators in the context of interval-valued
intuitionistic fuzzy values, which has as main characteristic by the best 𝕃∗
-representation of the usual
OWA and WA operators. Therefore, when applied to the diagonal elements these new operators have
the same behaviour as the OWA and WA. This paper also extended the notion of interval
representations introduced in [54] for 𝕃∗
-representations, and has introduced a new notion of inclusion
for 𝕃∗
-values which is based in a notion of membership. Besides, we introduced a new total order for
𝕃∗
-values and provide new extensions of the OWA operator for 𝕃 and 𝐿∗
-values.
Table 10. Comparing based on the Table 9.
We have shown the validity of our theoretical develpments by means of an illustrative decision-making
example. In [32] was introduced an interval-valued Atanassov’s intuitionistic extension of OWA’s where
the weights are assigned by decreasingly ordering the inputs with respect to an admissible order. The
problem with this OWA is that in general it is not increasing with respect to the admissible order. So, as
future work we intend to investigate OWAs on 𝕃∗
which are increasing with respect to a fixed admissible
order. In addition, based on [17], we will use such OWAs in a method to select the most important
vertice of an Interval-Valued Intuitionistic Fuzzy Graph [7].
Acknowledgment
This work was supported by the Brazilian funding agency CNPq (Brazilian Research Council) under
Project: 311429/2020-3.
References
da Cruz Asmus, T., Dimuro, G. P., Bedregal, B., Sanz, J. A., Pereira Jr, S., & Bustince, H. (2020). General
interval-valued overlap functions and interval-valued overlap indices. Information sciences, 527, 27-50.
Atanassov, K. (2016). Intuitionistic fuzzy sets. International journal bioautomation, 20, 1.
Atanassov, K. T. (1999). Interval valued intuitionistic fuzzy sets. In Intuitionistic fuzzy sets (pp. 139-177).
Physica, Heidelberg.
Atanassov, K. T. (1994). Operators over interval valued intuitionistic fuzzy sets. Fuzzy sets and
systems, 64(2), 159-174.
Atanassov, K. T. (1999). Intuitionistic fuzzy sets. In Intuitionistic fuzzy sets (pp. 1-137). Physica,
Heidelberg.
Atanassov, K., & Rangasamy, P. (2018). Intuitionistic fuzzy sets and interval valued intuitionistic fuzzy
sets. Advanced studies in contemporary mathematics, 28(2), 167-176.
Nayagam, V. L. G., Muralikrishnan, S., & Sivaraman, G. (2011). Multi-criteria decision-making method
based on interval-valued intuitionistic fuzzy sets. Expert systems with applications, 38(3), 1464-1467.
Bedregal, B. C. (2010). On interval fuzzy negations. Fuzzy sets and systems, 161(17), 2290-2313.
Bedregal, B., Beliakov, G., Bustince, H., Calvo, T., Mesiar, R., & Paternain, D. (2012). A class of fuzzy
multisets with a fixed number of memberships. Information sciences, 189, 1-17.
Bedregal, B. C., Dimuro, G. P., Santiago, R. H. N., & Reiser, R. H. S. (2010). On interval fuzzy S-
implications. Information sciences, 180(8), 1373-1389.
Bedregal, B., & Santiago, R. (2013). Some continuity notions for interval functions and
representation. Computational and applied mathematics, 32(3), 435-446.
Bedregal, B. C., & Santiago, R. H. N. (2013). Interval representations, Łukasiewicz implicators and
Smets–Magrez axioms. Information sciences, 221, 192-200.
𝐑𝟑 𝐀𝟐 𝐀𝟑 𝐀𝟓
A 2 – 5 1
A 3 3 – 1
A 5 7 5 –

259
An
interval-valued
atanassov's
intuitionistic
fuzzy
multi-attribute
group
decision
making
method
based
on
the
best
representation
of
the
WA
and
OWA
operators
Bedregal, B. R. C., & Takahashi, A. (2005, May). Interval t-norms as interval representations of t-norms.
The 14th IEEE international conference on fuzzy systems, 2005. FUZZ'05. (pp. 909-914). IEEE.
Bedregal, B. R., Dimuro, G. P., Reiser, R. H. S., Carvalho, J. P., Dubois, D., Kaymak, U., & da Costa
Sousa, J. M. (2009). An approach to interval-valued r-implications and automorphisms.
In IFSA/EUSFLAT Conf. (pp. 1-6). Lisbon, Portugal.
Beliakov, G., Bustince, H., Goswami, D. P., Mukherjee, U. K., & Pal, N. R. (2011). On averaging
operators for Atanassov’s intuitionistic fuzzy sets. Information sciences, 181(6), 1116-1124.
Beliakov, G., Pradera, A., & Calvo, T. (2007). Aggregation functions: A guide for practitioners (Vol. 221).
Heidelberg: Springer.
Bello Lara, R., González Espinosa, S., Martín Ravelo, A., & Leyva Vázquez, M. Y. (2015). Model for
static analysis in fuzzy graphs based on composite indicators of centrality. Revista cubana de ciencias
informáticas, 9(2), 52-65. (In Spanish). http://scielo.sld.cu/scielo.php?pid=S2227-
18992015000200004&script=sci_arttext&tlng=en
Bustince Sola, H., Barrenechea Tartas, E., Pagola Barrio, M., Fernández Fernández, F. J., Xu, Z.,
Bedregal, B., ... & Baets, B. D. (2016). A historical account of types of fuzzy sets and their
relationships. IEEE transactions on fuzzy systems, 24(1). https://hdl.handle.net/2454/32795
Bustince, H., Barrenechea, E., & Pagola, M. (2008). Generation of interval‐valued fuzzy and atanassov's
intuitionistic fuzzy connectives from fuzzy connectives and from Kα operators: Laws for conjunctions
and disjunctions, amplitude. International journal of intelligent systems, 23(6), 680-714.
Bustince, H., Bedregal, B., Campion, M. J., Da Silva, I., Fernandez, J., Indurain, E., ... & Santiago, R. H.
N. (2020). Aggregation of individual rankings through fusion functions: criticism and optimality
analysis. IEEE transactions on fuzzy systems. DOI: 10.1109/TFUZZ.2020.3042611
Bustince, H., & Burillo, P. (1995). Correlation of interval-valued intuitionistic fuzzy sets. Fuzzy sets and
systems, 74(2), 237-244.
Bustince, H., Fernández, J., Kolesárová, A., & Mesiar, R. (2013). Generation of linear orders for intervals
by means of aggregation functions. Fuzzy sets and systems, 220, 69-77.
Bustince, H., Galar, M., Bedregal, B., Kolesarova, A., & Mesiar, R. (2013). A new approach to interval-
valued Choquet integrals and the problem of ordering in interval-valued fuzzy set applications. IEEE
transactions on fuzzy systems, 21(6), 1150-1162.
Chiclana, F., Herrera-Viedma, E., Alonso, S., Alberto, R., & Pereira, M. (2008). Preferences and
consistency issues in group decision making. In Fuzzy sets and their extensions: representation, aggregation
and models (pp. 219-237). Springer, Berlin, Heidelberg.
da Costa, C. G., Bedregal, B. C., & Neto, A. D. D. (2011). Relating De Morgan triples with Atanassov’s
intuitionistic De Morgan triples via automorphisms. International journal of approximate reasoning, 52(4),
473-487.
Deschrijver, G., Cornelis, C., & Kerre, E. E. (2004). On the representation of intuitionistic fuzzy t-norms
and t-conorms. IEEE transactions on fuzzy systems, 12(1), 45-61.
Chen, S. M., & Tan, J. M. (1994). Handling multicriteria fuzzy decision-making problems based on
vague set theory. Fuzzy sets and systems, 67(2), 163-172.
da Silva, I. A., Bedregal, B., da Costa, C. G., Palmeira, E., & da Rocha, M. P. (2015). Pseudo-uninorms
and Atanassov’s intuitionistic pseudo-uninorms. Journal of intelligent & fuzzy systems, 29(1), 267-281.
da Silva, I. A., Bedregal, B., Santiago, R. H., & Neto, A. D. (2012). A new method for interval-valued
intuitionistic group decision making. In Segundo congresso brasileiro de sistemas fuzzy (pp. 282-294).
Da Silva, I. A., Bedregal, B., & Santiago, R. H. N. (2016). On admissible total orders for interval-valued
intuitionistic fuzzy membership degrees. Fuzzy information and engineering, 8(2), 169-182.
De Miguel, L., Bustince, H., Fernández, J., Induráin, E., Kolesárová, A., & Mesiar, R. (2016).
Construction of admissible linear orders for interval-valued Atanassov intuitionistic fuzzy sets with
an application to decision making. Information fusion, 27, 189-197.
De Miguel, L., Bustince, H., Pekala, B., Bentkowska, U., Da Silva, I., Bedregal, B., ... & Ochoa, G. (2016).
Interval-valued Atanassov intuitionistic OWA aggregations using admissible linear orders and their
application to decision making. IEEE transactions on fuzzy systems, 24(6), 1586-1597.
Deschrijver, G., & Kerre, E. E. (2003). On the relationship between some extensions of fuzzy set
theory. Fuzzy sets and systems, 133(2), 227-235.

260
Da
Silva
et
al.
|J.
Fuzzy.
Ext.
Appl.
2(3)
(2021)
239-261
Dimuro, G. P., Bedregal, B. C., Santiago, R. H. N., & Reiser, R. H. S. (2011). Interval additive
generators of interval t-norms and interval t-conorms. Information sciences, 181(18), 3898-3916.
Goguen, J. A. (1967). L-fuzzy sets. Journal of mathematical analysis and applications, 18(1), 145-174.
Grattan‐Guinness, I. (1976). Fuzzy membership mapped onto intervals and many‐valued
quantities. Mathematical logic quarterly, 22(1), 149-160.
Hickey, T., Ju, Q., & Van Emden, M. H. (2001). Interval arithmetic: From principles to
implementation. Journal of the ACM (JACM), 48(5), 1038-1068.
Hong, D. H., & Choi, C. H. (2000). Multicriteria fuzzy decision-making problems based on vague
set theory. Fuzzy sets and systems, 114(1), 103-113.
Kim, S. H., & Han, C. H. (1999). An interactive procedure for multi-attribute group decision making
with incomplete information. Computers & operations research, 26(8), 755-772.
Jahn, K. U. (1975). Intervall‐wertige Mengen. Mathematische nachrichten, 68(1), 115-132.
Lee, W. (2009, August). A novel method for ranking interval-valued intuitionistic fuzzy numbers
and its application to decision making. 2009 international conference on intelligent human-machine
systems and cybernetics (Vol. 2, pp. 282-285). IEEE.
Mesiar, R., & Komorníková, M. (2010). Aggregation functions on bounded posets. In 35 years of fuzzy
set theory (pp. 3-17). Springer, Berlin, Heidelberg.
Milfont, T., Mezzomo, I., Bedregal, B., Mansilla, E., & Bustince, H. (2021). Aggregation functions on
n-dimensional ordered vectors equipped with an admissible order and an application in multi-
criteria group decision-making. https://arxiv.org/abs/2101.12030
Mitchell, H. B. (2004). An intuitionistic OWA operator. International journal of uncertainty, fuzziness
and knowledge-based systems, 12(06), 843-860.
Moore, R. E. (1979). Methods and applications of interval analysis. Society for Industrial and Applied
Mathematics.
Paternain, D., Jurio, A., Barrenechea, E., Bustince, H., Bedregal, B., & Szmidt, E. (2012). An
alternative to fuzzy methods in decision-making problems. Expert systems with applications, 39(9),
7729-7735.
Pedrycz, W., Ekel, P., & Parreiras, R. (2011). Fuzzy multicriteria decision-making: models, methods and
applications. John Wiley & Sons.
Reiser, R. H. S., & Bedregal, B. (2014). K-operators: An approach to the generation of interval-valued
fuzzy implications from fuzzy implications and vice versa. Information sciences, 257, 286-300.
Reiser, R. H. S., Bedregal, B. C., & Dos Reis, G. A. A. (2014). Interval-valued fuzzy coimplications
and related dual interval-valued conjugate functions. Journal of computer and system sciences, 80(2),
410-425.
Reiser, R. H. S., Bedregal, B., & Visintin, L. (2013). Index, expressions and properties of interval-
valued intuitionistic fuzzy implications. TEMA (São Carlos), 14, 193-208.
Reiser, R. H. S., & Bedregal, B. (2017). Correlation in interval-valued Atanassov’s intuitionistic fuzzy
sets—conjugate and negation operators. International journal of uncertainty, fuzziness and knowledge-
based systems, 25(05), 787-819.
Sambuc, R. (1975) Φ-fuzzy functions: Application to diagnostic aid in thyroid pathology (Phd
Thesis, University of Marseille, France). (In French).
https://www.scirp.org/(S(i43dyn45teexjx455qlt3d2q))/reference/ReferencesPapers.aspx?ReferenceID=1235361
Santana, F., Bedregal, B., Viana, P., & Bustince, H. (2020). On admissible orders over closed
subintervals of [0, 1]. Fuzzy sets and systems, 399, 44-54.
Santiago, R. H. N., Bedregal, B. R. C., & Acióly, B. M. (2006). Formal aspects of correctness and
optimality of interval computations. Formal aspects of computing, 18(2), 231-243.
Santiago, R., Bedregal, B., Dimuro, G. P., Fernandez, J., Bustince, H., & Fardoun, H. M. (2021).
Abstract homogeneous functions and Consistently influenced/disturbed multi-expert decision
making. IEEE Transactions on fuzzy systems. DOI: 10.1109/TFUZZ.2021.3117438
Shang, Y. G., Yuan, X. H., & Lee, E. S. (2010). The n-dimensional fuzzy sets and Zadeh fuzzy sets
based on the finite valued fuzzy sets. Computers & mathematics with applications, 60(3), 442-463.
Sirbiladze, G. (2021). New view of fuzzy aggregations. part I: general information structure for
decision-making models. Journal of fuzzy extension and applications, 2(2), 130-143. DOI:
10.22105/jfea.2021.275084.1080

261
An
interval-valued
atanassov's
intuitionistic
fuzzy
multi-attribute
group
decision
making
method
based
on
the
best
representation
of
the
WA
and
OWA
operators
Tan, C. (2011). A multi-criteria interval-valued intuitionistic fuzzy group decision making with Choquet
integral-based TOPSIS. Expert systems with applications, 38(4), 3023-3033.
Tan, C., & Chen, X. (2013). Interval-valued intuitionistic fuzzy multicriteria group decision making based
on VIKOR and Choquet integral. Journal of applied mathematics. https://doi.org/10.1155/2013/656879
Tan, C. Y., Deng, H., & Cao, Y. (2010). A multi-criteria group decision making procedure using interval-
valued intuitionistic fuzzy sets. Journal of computational information systems, 6(3), 855-863.
Tizhoosh, H. R. (2008). Interval-valued versus intuitionistic fuzzy sets: Isomorphism versus
semantics. Pattern recognition, 41(5), 1812-1813.
Wan, S., & Dong, J. (2014). A possibility degree method for interval-valued intuitionistic fuzzy multi-
attribute group decision making. Journal of computer and system sciences, 80(1), 237-256.
Wan, S., & Dong, J. (2020). Decision making theories and methods based on interval-valued intuitionistic fuzzy
sets. Springer Nature, Singapore Pte Ltd.
Wang, Z., Li, K. W., & Wang, W. (2009). An approach to multiattribute decision making with interval-
valued intuitionistic fuzzy assessments and incomplete weights. Information sciences, 179(17), 3026-3040.
Wei, G. (2010). Some induced geometric aggregation operators with intuitionistic fuzzy information and
their application to group decision making. Applied soft computing, 10(2), 423-431.
Ze-Shui, X. (2007). Methods for aggregating interval-valued intuitionistic fuzzy information and their
application to decision making [J]. Control and decision, 2, 019.
Xu, Z. S., & Jian, C. H. E. N. (2007). Approach to group decision making based on interval-valued
intuitionistic judgment matrices. Systems engineering-theory & practice, 27(4), 126-133.
Xu, Z., & Yager, R. R. (2006). Some geometric aggregation operators based on intuitionistic fuzzy
sets. International journal of general systems, 35(4), 417-433.
YagerR, P. (1988). On ordered weighted averaging operators in multicritefia decision making. IEEE
transactions on systems, man and cybernetics, 18, 183-190.
Yager, R. R. (2009). OWA aggregation of intuitionistic fuzzy sets. International journal of general
systems, 38(6), 617-641.
Zadeh, L. A. (1965). Fuzzy sets. Information and control, 8(3), 338-353.
Zadeh, L. A. (1975). The concept of a linguistic variable and its application to approximate reasoning—
I. Information sciences, 8(3), 199-249.

An interval-valued atanassov’s intuitionistic fuzzy multi-attribute group decision making method based on the best representation of the WA and OWA operators

More Related Content

Similar to An interval-valued atanassov’s intuitionistic fuzzy multi-attribute group decision making method based on the best representation of the WA and OWA operators (20)

More from Journal of Fuzzy Extension and Applications (20)

Recently uploaded (20)

An interval-valued atanassov’s intuitionistic fuzzy multi-attribute group decision making method based on the best representation of the WA and OWA operators