Mathematical Problems in Engineering.

Multiple criteria decision making approach with multi-valued neutrosophic

linguistic normalized weighted Bonferroni mean Hamacher operator

Bao-lin Li1，2, Juan-ru Wang1 , Li-hua Yang2 , Xue-tao Li2

1.
School of Management, Northwestern Polytechnical University, Xi’an 710072, China;
2.
School of Economics and Management, Hubei University of Automotive Technology,
Shiyan 442002, China
Corresponding author: Juan-ru Wang
E-mail: wjuanru@nwpu.edu.cn

Abstract The neutrosophic set and linguistic term set are widely applied in recently years. Motivated
by the advantages of them, we co mbine the mult i-valued neutrosophic set and linguistic set and define
the concept of the mult i-valued neutrosophic linguistic set (MVNLS). Furthermore, Hamacher
operation is an extension of the Algebraic and Einstein operation. Additionally, the normalized
weighted Bonferroni mean (NW BM) operator can consider the weight of each argument as well as
capture the interrelationship of different arguments . Therefore, the co mbination of NWBM operator
and Hamacher operation is more valuable and agile. Firstly, M VNLS and mult i-valued neutrosophic
linguistic number (M VNLN) are defined, then some new operational ru les of MVNLNs on account of
Hamacher operations are developed, and the comparison functions for MVNLNs are given. Secondly,
mu lti-valued neutrosophic linguistic normalized weighted Bonferroni mean Hamacher o perator
(M VNLNW BMH) is proposed, and a number of expected characteristics of new operator are
investigated. Meanwhile, some special cases of different parameters p ,q and  are analy zed. Th ird ly,
the approach utilizing the MVNLNW BMH operator is introduced to manage multip le criteria decision
making issue (MCDM) in mu lt i-valued neutrosophic linguistic environ ment. Ultimately, a practical
example is presented and a comparative analysis is carried out, which validate the effectiveness and
generalization of the novel approach.
Keywords Multi-valued neutrosophic linguistic, NWBM, Hamacher, Multiple criteria decision making

1 Introduction
In real world, due to the complexity of decision informat ion, the fuzzy theory has attracted
widespread attentions and has been developed in various fields. Zaheh [1] firstly proposed the notion of
fuzzy sets (FSs). Then, Atanassov [2] introduced the intuitionistic fuzzy sets (IFSs), which overco me
the weakness of non-membership degrees. Subsequently, in order to address the
hesitation degree of decision-makers, Torra [3] defined hesitant fuzzy sets (HFSs). Fuzzy set theory has
gained well pro moted, but it still cannot manage the inconsistent and indeterminate information. Under
this circu mstance, Smarandache [4] proposed Neutrosophic Sets (NSs), whose indeterminacy degree is
independent on both true and false membership. NS is an extension of IFS, and makes decision-makers
express their preference more accurately, so some achievements on NSs and its extensions have been
undertaken. Some various concepts of different NSs are defined. For examp le, Smarandache [54] and
Wang et al. [5] introduced single-valued neutrosophic sets (SVNs) to facilitate its application. Ye [6]
pointed out the concept of simp lified neutrosopic sets (SNSs).Wang [7] developed the concept of
interval neutrosopic sets (INSs). However, under certain conditions, the decision makers likely give
different evaluation numbers for expressing their hesitant. Subsequently, the definition of single-valued
neutrosophic hesitant fuzzy sets (SVNHFSs) was firstly proposed by Ye [8] in 2014, then Wang [9]
also proposed mult i-valued neutrosophic sets (MVNSs) in 2015. Actually, the notions of SVNHFSs
and MVNSs are equal. For simplicity, we adapt the term of MVNSs in this paper.

1
Mathematical Problems in Engineering.

On the other hand, the aggregation operators, comparison method for Neutrosophic numbers are also
been studied. For SVNSs, Liu [10] employed NWBM operator to solve mult iple criteria problem in
single-valued neutrosophic environment. Ye [11] gave the definitions of cross-entropy and correlation
coefficient. Fo r INSs, Zhang [12] developed some aggregation operators. Liu [13] not only provided
the definition of interval neutrosophic hesitant fuzzy sets (INHFSs), but also discussed the generalized
hybrid weighted average operator. Broumi and Smarandache [14-16] studied the correlation
coefficients, cosine similarity measure and some new operations. Ye [17] proposed similarity measures
between interval neutrosophic sets. For M VNSs, Ye [8] developed SVNHFWA and SVNHFW G
operators for MCDM problem. Peng [18, 19] extended power aggregation operators and defined some
outranking relat ions under MVNS environ ment. Ji et al. [20] analy zed a novel TODIM method for
MVNSs.
In real life, o wing to the amb iguity of decision makers’ thinking, people prefer to utilize linguistic
variables for describing their assessment value rather than the quantization value. Therefore, linguistic
variable has attracted widespread attention in the field of MCDM. The linguistic variab le was firstly
proposed by Zadeh [21] and applied fo r the fuzzy reasoning. After that, a series of works on it have
been made, Wang [22-24] presented a new approach in view of hesitant fuzzy linguistic informat ion.
Meng [25] developed linguistic hesitant fuzzy sets and studied hybrid weighted operator. Tian [26]
defined gray linguistic weighted Bonferroni mean operator for MCDM.
In order to indicate the true, indeterminate and false extents concerning a linguistic term, the NSs
and linguistic set (LS) are co mb ined. Several neutrosophic linguistic sets and its corresponding
operator are defined, for examp le single valued o r simp lified neutrosophic linguistic sets and trapezoid
linguistic sets [27-30], interval neutrosophic certain or uncertain linguistic sets [31-33].However, due
to the hesitancy of people’s thinking, the true of a linguistic term may be given several values, and the
case is similar to the false and indeterminate extents. The existing literatures don’t consider this
perspective. Therefore, the mult i-valued neutrosophic linguistic set (MVNLS) and mult i-valued
neutrosophic linguistic number (M VNLN) in this article are proposed in order to better express the
information.
Aggregation operator which can fuse multiple arguments into a single comprehensive value is an
important tool for MCDM p roblem. Many researchers have developed some efficient operators [34-41],
for instance, the weighted geometric average (W GA ) or averag ing (WA) operator, prioritized
aggregation (PA) operator, Maclaurin symmetric mean and Bonferroni mean (BM) operator. BM
operator was originally defined by Bonferroni [42], which has attracted widespread attentions because
of its characteristics of capture interrelat ionship among arguments. Some ach ievements have been
made on it [43-49].In order to aggregate neutrosophic linguistic informat ion, some researches on
aggregation operators under neutrosophic linguistic and neutrosophic uncertain linguistic environ ments
are also been applied [27-33,50].Until to now, BM and NWBM fail to accommodate aggregation
informat ion for mult i-valued neutrosophic linguistic environ ment. Motivated by this limitat ion, we will
extend the NWBM operator to MVNLS in this article.
T-norms and t-conorms are two functions that satisfy certain conditions respectively. The
Archimedean t-conorms and t-norms are well-known, which include algebraic, Einstein and Hamacher,
Hamacher operation is an extension of algebraic and Einstein. Generally, the algebraic operators are
commonly, there are also a few aggregation operations based on Einstein operations. Due to Hamacher
operator is more general, Liu [51, 52] discussed the Hamacher operational rules. So far, there is no
research for M VNLS based on Hamacher operations. Since it is better for M VNLS to depict the actual
situation, NWBM operator can capture the interrelationship among arguments, and Hamacher
operations are more general, it is of great meaning to study the NWBM Hamacher operators under
multi-valued neutrosophic linguistic environment for MCDM problems.
The main purposes of the paper are presented in the following:
1. To be better exp ress people’s hesitant, comb ining the M VNS and LS, we give the notions of
MVNLS and M VNLN, besides, the score, accuracy and certainty functions are also investigated to
compare MVNLNs.

2
Mathematical Problems in Engineering.

2. Due to the generalizat ion of Hamacher operational rules, we define new operations of M VNLNs
based on Hamacher operational rules, and discuss their operational relation s.
3. The NWBM considering the interrelationship of different argu ments has gained widespread
concerns, we extend NWBM operator to M VNLN environ ment, the M VNLNW BMH operator is
defined, moreover, some desirable characteristics are also studied.
4. In order to verify the effectiveness, an examp le for M CDM problem utilizing M VNLNWBM H
operator is illustrated and conduct a comparative analysis. We also analyze the influence s of
different parameter values for the final outcomes, the results demonstrate the operator proposed is
more general and flexible.
The article is arranged in this way. In Section 2, we review a nu mber o f notions and operations for
MVNS, LS, NW BM operator and Hamacher. In Sect ion 3, we propose the definitions of MVNLS and
MVNLN, and develop the operations of M VNLNs on the basis of Hamacher t -conorms and t-norms.
Meanwhile, the A lgebraic as well as Einstein for M VNLNs are also presented, which are special cases
of Hamacher operation. Moreover, the comparison method of M VNLNs is also defined. In Section 4,
we propose the M VNLNW BMH operator and investigate its properties. Furthe rmo re, when
corresponding parameters are assigned different values, the special examp les are also discussed. In
Section 5, we establish the MCDM procedure on account of the proposed aggregation operators with
MVNLS in formation. Section 6 presents a concrete example, as well as a comparison analysis is
provided to show the practicability utilizing our method. Finally, in Section 7, some results are
presented.

2 Preliminaries
Some notions and operation are introduced in this section, which will be useful in the latter
analysis.
2.1 Linguistic term sets
Suppose that S  s1 ,s2 , ,s  is an ordered and finite linguistic set, in which s j denotes a
linguistic variable value and  is an odd value. When  is equal to seven, the corresponding linguistic
set are provided in the following:
S  s1 ,s2 ,s3 ,s 4 ,s5 ,s6 ,s7   extremely poor,very poor,poor,medium,good,very good,extremely good .
In order to avoid the linguistic information loss, the set above is expanded, that is a contiguous set,
S  s   R  .
Definiti on 1 [53] Let si and s j be any two linguistic variables, the corresponding operations are
presented:

(1) si  s  i ,  0;
(2)si  s j  si  j ;
(3)si  s j  si  j ;

(4) si 

 si  .

2.2 Multi-valued neutrosophic sets

Definition 2 [8,9] Suppose that X is a collection of objects, MVNSs A on X is defined by

A   x ,T (x ),I (x ),F (x )
A A A 
x X ,

    
Where TA(x )     TA(x ) ,I A(x )     I A(x ) , FA(x )     FA(x ) , 
TA(x ),I A(x ),and FA(x ) are three collections of crisp numbers belonging to 0,1 , representing the
probable true-membership degree, indeterminacy-membership degree and falsity- membership degree,

3
Mathematical Problems in Engineering.

where x in X belonging to A , respectively, satisfying these conditions 0   , ,  1, and

0  sup TA(x )  sup I A(x )  sup FA(x )  3. If there is only one element in X , A is indicated by the
three tuple A  TA(x ),I A(x ),FA(x ) , that is known as a mu lti-valued neutrosophic number
(M VNN).Generally, M VNSs is considered as the generalizations of the other sets, such as FSs, IFSs,
HFSs, DHFs, and SVNSs.

2.3 Normalized weighted Bonferroni mean

Definiti on 3 [42] Let p ,q  0,as well as ai(i  1,2, n ) be a set of nonnegative

values, then the BM is defined as
1

  p q
 
n
 1 
 n(n  1) i 
BM p ,q(a1 ,a2 , ,an )  aip ajq
,j 1,

 i j 
Definiti on 4 [45]Let p ,q  0,and ai(i  1,2, n ) be a set of nonnegative values, and
the corresponding NWBM can be expressed as below:
1

 n ww  p q
NWBM p ,q
(a1 ,a2 , ,an )    1w (ai  aj )
i j p q
 i ,j 1, i 
 i j 
Where w  ( w1 , w2 , , wn ) represents the corresponding weighted vector
n
of ai i  1,2, , n  , satisfying w i  0 and  w i  1 . The weight vector can be
i 1
given by decision-makers in real problem.
Obviously, the NWBM operator possesses a few characteristic s such as
commutativity, reducibility, monotonicity, boundedness, and idempotency.

2.4 Hamacher operations

We know aggregation operator is given in accordance with different t-norms and t-
conorms, there are some exceptional circumstances listed in the following:
(1) Algebraic t-norm and t-conorm
a  b  ab,a  b  a  b  ab;
(2) Einstein t-norm and t-conorm
ab a b
a b  ,a  b  ;
1  1  a   1  b  1  ab

(3) Hamacher t-norm and t-conorm

ab a  b  ab  (1   )ab
a b  ,a  b  ,  0
  1    a  b  ab  1  (1   )ab

4
Mathematical Problems in Engineering.

In special, when   1,  2, the Algebraic and Einstein operations are the
simplifications of Hamacher t-norm and t-conorm.

3 Multi-valued neutrosophic linguistic set

3.1 MVNLS and its Hamacher operations
Definition 5 Let X be a set of points, an MVNLS A in X is defined as follows:

A   x ,s ( x ) 
, TA(x ),I A(x ),FA(x )  x  X ,
  
 
Where s(x )  S , TA(x )     TA(x ) ,I A(x )     I A(x ) , FA(x )     FA(x ) ,    
TA(x ),I A(x ), and FA(x ) are three sets of crisp values in 0,1 , denoting three degrees of
x in X belonging to s(x ), , that are true, indeterminacy and falsity, satisfying these conditions
0   , ,  1,and 0  sup TA(x )  sup I A(x )  sup FA(x )  1.

Definiti on 6 Let A   x ,s ( x )   

, TA(x ),I A(x ),FA(x )  x  X be an MVNLS, supposing there is


only one element in X , then tuple s(x ),TA(x ),I A(x ),FA(x ) is depicted as a mu lt i-valued
neutrosophic linguistic number (MVNLN).For simplicity, the MVNLN can also be represented as

A   s ,T (x ),I (x ),F (x )

( x ) A A A
x X 
Definiti on 7 Let a1  s(a1 ), T(a1 ),I(a1 ),F(a1 )   
and a2  s(a2 ), T(a2 ),I(a2 ),F(a2 )  be two
MVNLNs, and   0,then the operations of MVNLNs can be defined on the basis of Hamacher
operations.
(1) a1  a2
 s(a ) (a ),
1 2

   1   2   1 2  (1   ) 1 2 
  ,

 1 T (a1 ), 2 T (a2 )
 1  (1   ) 1 2 
 1 2 
 ,
   (1   )(1   2  1 2 )
1 I (a1 ), 2 I (a2 )

 12  
   ;
   (1   )(1  2  12 ) 
1 F (a1 ),2 F (a2 )

(2) a1  a2
 s(a )(a ),
1 2

   1 2 
  ,
   (1   )( 1   2   1 2 )
 1 T (a1 ), 2 T (a2 )

 1   2  1 2  (1   )1 2 
 ,
1 I (a1 ), 2 I (a2 )
 1  (1   )1 2 
1  2  12  (1   )12  
   ;
1 F (a1 ),2 F (a2 )
 1  (1   )12  

5
Mathematical Problems in Engineering.

(3)  a1  s (a ),
1

  (1  (  1) 1 )  (1   1 ) 
   
,
  1 T (a1 ) 
(1  (  1) 1 )  (  1)(1   1 ) 

 1 
1 I (a1 )    
,
(1  (  1)(1  1 ))  (  1)1 
 1  
1 F (a1 )    
 ;
(1  (  1)(1  1 ))  (  1)1  

(4) a1  s  (a ),
1

   1 
   
,
  1 T (a1 ) 
(1  (  1)(1   1 ))  (  1) 1 

 (1  (  1)1 )  (1  1 ) 
1 I (a1 )    
,
(1  (  1)1 )  (  1)(1  1 ) 

 (1  (  1)1 )  (1  1 )  
1 F (a1 )    
 .
(1  (  1)1 )  (  1)(1  1 )  
If   1,then the operations based on Hamacher operational ru les in Definit ion 7 will simp lified to
the Algebraic operational rules in the following:
(5) a1  a2
 s(a ) (a ),
1 2

  1 T (a1 ), 2 T (a2 )        ,

1 2 1 2

1 I (a1 ), 2 I (a2 )    ,

1 2

1 F (a1 ),2 F (a2 )    ;

1 2

(6) a1  a2
 s(a )(a ),
1 2

  1 T (a1 ), 2 T (a2 )    ,1 2

1 I (a1 ), 2 I (a2 )        ,

1 2 1 2

1 F (a1 ),2 F (a2 )        ;

1 2 1 2

(7)  a1  s (a ),
1

  1 T (a1 ) 1  (1   )  , 1


1 I (a1 )   ,

1

1 F (a1 )   ;
1


6
Mathematical Problems in Engineering.

(8) a1  s  (a ),
1

  1 T (a1 )   ,
1


1 I (a1 ) 1  (1   )  , 1


1 F (a1 ) 1  (1   )  . 1


Supposing T (a1 ),I(a1 ),F (a1 ),T (a2 ),I(a2 ), and F (a 2 ) contain only one value, then the operations
defined above can be reduced to the operations of SVNLNs based on Algebraic operations proposed by
Ye [27].
If   2,then the operations based on Hamacher operational ru les in Defin ition 7 will simp lified to
the Einstein operations of MVNLNs presented below:
(9) a1  a2
 s(a ) (a ),
1 2

   1   2 
  ,
1   1 2 
 1 T (a1 ), 2 T (a2 )

 1 2 
 ,
 2  1   2  1 2 
1 I (a1 ), 2 I (a2 )

 12  
  ;
 2  1  2  12  
1 F (a1 ),2 F (a2 )

(10) a1  a2
 s(a )(a ),
1 2

   1 2 
  ,
 2   1   2   1 2 
 1 T (a1 ), 2 T (a2 )

 1   2 
 ,
1  1 2 
1 I (a1 ), 2 I (a2 )

 1  2  
  ;
1  12  
1 F (a1 ),2 F (a2 )

(11)  a1  s (a ),
1

 (1   1 )  (1   1 ) 
   
,
  1 T (a1 ) 
 (1   1 )  (1   1 ) 
 21 
1 I (a1 )    
,
(2  1 )  1 
 21  
1 F (a1 )    
 ;
(2  1 )  1  

7
Mathematical Problems in Engineering.

(12) a1  s  (a ),
1

  2 1 
   
,
  1 T (a1 ) 
 (2   1 )   1 
(1  1 )  (1  1 ) 
1 I (a1 )    
,
(1  1 )  (1  1 ) 
(1  1 )  (1  1 )  
1 F (a1 )    
 .
(1  1 )  (1  1 )  
Supposing T (a1 ),I(a1 ),F (a1 ),T (a2 ),I(a2 ),and F (a 2 ) contain only one value, then the operations
defined above can be reduced to the operations of SVNLNs based on Einstein operations.
Theorem 1 Let a1  s(a ),T(a1 ),I(a1 ),F(a1 ) , a2  s(a ),T(a2 ),I(a2 ),F(a2 ) , and
2
1

a3  s(a ),T(a3 ),I(a3 ),F(a3 ) be any three M VNLNs, and  ,1 ,2  0, then the properties
3

below are correct:

(1) a1  a2  a2  a1;
(2) a1  a2  a2  a1;
(3)  a1  a2   a1  a2;
(4) 1a1  2a1  (1  2 )a1;
(5) a11  a12  a11  2 ;

(6) a1  a2  a1  a2  ;



(7) a 1
 a2   a3  a1  a2  a3  ;
(8) a 1
 a2   a3  a1  a2  a3  .

Then, equation (4) will be proved as follows:

Proof (4) Since 1 ,2  0,

1a1  2a1  s 1(a1 ) 2(a1 ),

  (1  (  1) 1 )1  (1   1 )1 (1  (  1) 1 )2  (1   1 )2 (1  (  1) 1 )1  (1   1 )1 (1  (  1) 1 )2  (1   1 )2 (1  (  1) 1 )1  (1   1 )1 (1  (  1) 1 )2  (1   1 )2 
   
  
  
  
 (1   )  
 (1  (  1) 1 ) 1  (  1)(1   1 ) 1 (1  (  1) 1) 2  (  1)(1   1) 2 (1  (  1) 1) 1  (  1)(1   1) 1 (1  (  1) 1 ) 2  (  1)(1   1 ) 2 (1  (  1) 1 )1  (  1)(1   1)1 (1  (  1) 1)2  (  1)(1   1)2 
 1 T (a1 )  1 1 2 2 ,
  (1  (  1) 1 )  (1   1 ) (1  (  1) 1 )  (1   1 ) 
  1  (1   )  
  (1  (  1) 1 )1  (  1)(1   1 )1 (1  (  1) 1 )2  (  1)(1   1 )2 
 11 12 
  
 (1  (  1)(1  1 ))1  (  1)11 (1  (  1)(1  1))2  (  1)12 
1 I (a1 )  1 2 1 2 ,
   (1   )( 1 1 1 1
 1 1
 2 2
 1 1
 2
)
2 
 (1  (  1)(1  1 ))  (  1)1 (1  (  1)(1  1 ))  (  1)1 (1  (  1)(1  1 ))  (  1)1 (1  (  1)(1  1))  (  1)1 
 11 12 
  
 (1  (  1)(1  1 ))1  (  1)11 (1  (  1)(1  1 ))2  (  1)12 
1  F (a1 )  1 2 1 2 
   (1   )( 1 1 1 1
 1 1
 2 2
 1 1
 2
)
2 
 (1  (  1)(1  1 ))  (  1)1 (1  (  1)(1  1 ))  (  1)1 (1  (  1)(1  1))  (  1)1 (1  (  1)(1  1))  (  1)1 

 s(1  2 )(a1 ),
  (1  (  1) 1 ) 1 2  (1   1 ) 1 2
   

  ,
 1 T (a1 ) 1  2
 (  1)(1   1 )1  2 
 (1  (  1) 1 )
 11  2 
1 I (a1 )  1  2 ,
(1  (  1)(1  1 ))  (  1)11  2 
 11  2  
1  F (a1 )  1  2  
(1  (  1)(1  1 ))  (  1)11  2 

 (1  2 )a1

Therefore, equation (4) 1a1  2a1  (1  2 )a1 can be obtained.

8
Mathematical Problems in Engineering.

Similarly, the other equations in Theorem1 are easily certified in the light of Definition
7.
3.2 Comparison method
The sco re, accu racy , and certainty fun ct ions are impo rtant inde xes to ran k M VNLNs , an d its
corresponding definition is given below:

Definiti on 8 Let a  s(a ),T (a ),I(a ),F(a ) be an M VNLN, and the sco re, accuracy , and
certainty functions are achieved as below.

 1  1 1 
(1) E(a )  
      T (a ), I (a ), F (a )
(
3
) s(a )

 T (a ) I (a ) F (a ) 
 s 1   1    1  

     T (a ), I (a ), F (a )( 3
) (a )


T (a ) I (a ) F (a ) 

 1 
(2) H(a )  
    T (a ), F (a )
(   ) s(a )

 T (a ) F (a ) 
 s 1 

    T (a ), F (a )(  )(a )

T (a ) F (a ) 

 1 
(3) C(a )  
  T (a )
  s(a )

 T (a ) 
 s 1 

   T (a )  (a )

T (a ) 

Where T (a ),I (a ), and F (a ) are the numbers of the values in TA(x ),I A(x ), and FA(x ) respectively.
The linguistic variable s(a ) is important for an M VNLN. Therefore, the co mparison functions defined
above in Definition 8 are denoted as the linguistic variable. The bigger the truth degree T (a )
concerning the variable s(a ) is, meanwh ile, the smaller the indeterminacy degree I(a ) as well as the
false degree F(a ) concerning the linguistic variable s(a ) are, then the higher the MVNLN is. Relating
to the function of score, the greater      corresponding to s(a ) is, the higher the affirmat ive
statement is. Relating to the function of accuracy, the greater  minus  is, the certain the statement is.
Regarding to the function of certainty, the bigger  is, the certain the statement is.
Based on Definition 8, the comparison method between MVNLNs are obtained.
Definiti on 9 Supposing a1 and a2 are two MVNLNs, the compared approach is achieved in the
following:
(1) Supposing that E(a1 )  E(a2 ) , then a1 is greater than a2 , represented as a1 a2 ;
(2) Supposing that E(a1 )  E(a2 ) , and H (a1 )  H (a2 ) ,then a1 is greater than a2 , represented
as a1 a2 ;
(3) Supposing that E(a1 )  E(a2 ) , H (a1 )  H (a2 ) , and C(a1 )  C(a2 ) ,then a1 is greater than a2 ,
represented as a1 a2 ;
(4) Supposing that E(a1 )  E(a2 ) , H (a1 )  H (a2 ) , and C(a1 )  C(a2 ) ,then a1 equals a2 ,
represented as a1 a2 ;

9
Mathematical Problems in Engineering.

4 The multi-val ued neutrosophic linguistic normalized weighted Bonferroni mean Hamacher
operator
The NWBM operator can not only take into account the advantages of BM and WBM, but also has
the property of reducibility and idempotency. However, the NW BM operator has not been applied to
the cases where the input arguments are MVNLNs.

Definiti on 10 Let ai(i  1,2, ,n ) be a space of M VNLNs,

ai  s(ai ),T(ai ),I(ai ),F(ai ) , p ,q  0, and   1 ,2 , ,n  be the weighted vector
n
for ai ,i  0,1 and 
1
i
i  1.Then the operator of M VNLNWBM H is achieved as below, the

aggregation result is still an MVNLN.

1
 n  p q
 i j q 
MVNLNWBMH (a1 ,a2 , an )   (ai  aj )
p
 i ,j 1 1   
i
 i j 
According to the operational laws in Definition 7, the results are derived below:

10
Mathematical Problems in Engineering.

MVNLNWBMH (a1 ,a2 , an )  s 1 ,

  p q
 n  i j p 
  
 i ,j  1  1  i
 (ai ) q (aj ) 


i j 

  1 
   x y  p q 
    
  
 x    1 y 
 
  i T (ai ), j T (aj )  1 1
,
     p q   p q 
x y
   1     1  1      1  x  x  y 1 y  





 x     1 y 
    



 1 1

  h  p q  h  p q 
  1     1   1   


 g     1 h  
 g     1 h  
i I (ai ), j I (aj )  1 1 ,
  p q   p q 
h h
 1     1      1  1   


g     1 h  
 g     1 h  

 1 1

  u  p q  u  p q 
  1     1   1   


 v     1 u  
 v     1 u  
i F (ai ),j F (aj )  1 1 
  p q   p q 
u u
 1     1      1  1   


v     1 u  
 v     1 u 




 s 1 ,
  p q
 n  i j p 
  
 i ,j  1  1  i
 (ai ) q (aj ) 


i j 

  1 
   x  y  p q 
  i T (ai ), j T (aj )  ,
 1

   
1
 x   2  1 y p q    1 x  y p q 

      
 1

   
1
 g   2  1 h p q   g  h  p q 
i I (ai ), j I (aj )  1 ,
   
1
 g   2  1 h p q    1 g  h p q 
     
 1

   
1
 v   2  1 u p  q  v  u  p  q  
i F (ai ),j F (aj )  

1

   
1
 v   2  1 u p q    1 v  u p q
     
(1)
i j

   
n

     1  i      1  j 
p q 1  i
Where x    2  1 ip jq ,
, 1
i j
i j

11
Mathematical Problems in Engineering.

i j

 
n

     1  i      1  j 
p q
q 1  i
y   i  j p
,
i ,j  1
i j
i j

 
n

 1    1 i  1    1  j   (  1) 1  i  1   
p q p q 1  i
g  2
j
,
i ,j  1
i j
i j

 
n

 1    1 i  1    1  j   1  i  1   
p q p q 1  i
h  j
,
i ,j  1
i j
i j

 
n

 1    1 i  1    1 j   (  1) 1  i  1   
p q p q 1  i
v  2
j
.
i ,j  1
i j
i j

 
n

 1    1 i  1    1 j   1  i  1   
p q p q 1  i
u  j
,
i ,j  1
i j

Proof. According to the operational rules for MVNLNs, the results below can be gained
  
   i p 
ai p
 s p (a ),  ,
 
 1     1 1   i      1  i p
 i T (ai ) p
i
 
  


1     1 i  1  i   
p p
 
i I (ai )  ,
 
 1     1 i     1 1  i 
p p

 


1     1 i  1  i   
p p
 
i F (ai )  
 
 1     1 i     1 1  i 
p p

 
  
    jq 
aj q  s q (a ),  j T (aj )  ,
 
 1     1 1   j      1  j q
q
j
 
  


1     1  j  1   j   
q q
 
 j I (aj )  ,
 
 1     1  j     1 1   j 
q q

 


1     1 j  1  j   
q q
 
j F (aj )  
 
 1     1 j     1 1  j 
q q

 

12
Mathematical Problems in Engineering.

  
   i p  j q 
ai p  aj q  s  p (a ,  i T (ai ), j T (aj )  ,
i ) (aj ) 

      1  i       1  j     1  i p  j q
q p q
 
  

  
1     1 i  1     1  j  1  i  1   j   
p q p q
 
i I (ai ),j I (aj )  ,
  
 1     1 i  1     1  j     1 1  i  1   j  
p q p q

 

  
1     1 i  1     1 j  1  i  1  j   
p q p q
 
i F (ai ),j F (aj )  
  
 1     1 i  1     1 j    1 1  i  1  j  
p q p q

 

Firstly, we need to testify the mathematical formula below.

n i j

i ,j  1 1  i
(ai p  aj q )
i j

  i j i j

     
n n
 
  
     1  i     1  j   2  1  i p  j q i  i     1  i     1  j   i p  j q   
p q p q
1 1  i

  i ,j  1 ,j  1 
  i j i j 
 s n  ,  i T (ai ), j T (aj )  i j i j ,

   
 1  i  (ai ) (aj ) 
i j p q
 n 
 
n
       1  i   
     1  j   2  1  i p  j q    1       1  i      1  j   i p  j q   
p q 1  i p q 1  i

i ,j  1
i j

  ii ,jj1 i ,j  1 
  i  j 
 i j

1    1   
n
 1    1  j   1  i  1   
p q

p q 1  i

 i ,j  1
i j

 i j 
i I (ai ),j I (aj )  i j i j ,
 n
    
n
  1     1 i  1    1  j   ( 2  1) 1  i  1        1  1    1  i  1    1  j   1  i  1   
p q p q 1  i p q p q 1  i
j j 
 ii ,jj1 , 1
i j 
 i j 
 i j

1    1   
n

1    1 j   1  i  1    
p q

p q 1  i

 i ,j  1
i j

 i j 
i F (ai ),j F (aj )  i j i j  .
 n
    
n
  1     1 i  1    1 j   ( 2  1) 1  i  1        1  1    1 i  1    1  j   1  i  1   
p q p q 1  i p q p q 1  i
j j 
 ii ,jj1 , 1
i j 
 i j 

(2)
The mathematical induction on n is adopt to prove Eq.(2)

(1) Supposing n  2,the equation below is obtained.

2 i j 12 

i j
,  1 1  i
(ai p  aj q ) 
1  1
(a1p  a2q )  2 1 (a2 p  a1q )
1  2
i j

  x 1x 2  y 1y 2 
 s  ,  ,
 x 1x 2     1 y 1y 2 

1 2
 (a1 ) (a2 ) 2 1  p (a2 ) q (a1 )
p q
  i T (ai ), j T (aj )
1  1 1  2 
  m1m 2 
 ,
 1 2 
i I (ai ), j I (aj )
z z    1 m 1m 2 
 v 1v 2  
 
 1 2 
i F (ai ),j F (aj )
u u    1 v 1v 2  
12
1  1

 
Where x   1     1   1p  2q  ,
 
     1  1        1  2      1  1p  2q
1 p q
 
 

13
Mathematical Problems in Engineering.

21
1  2

 
 2 p  1q
x 2   1     1   ,
p q 
    1  2       1  1      1  2  1 
p q

 
12
1  1

 
 1p  2q
y 1   1  

,
    1  1       1  2      1  1p  2q
p q
 
 
21
1  2

 
 2 p  1q
y 2   1   ,
p q 
    1  2       1  1     1  2  1 
p q

 
12
1  1


1     1    1    1    1   
 1 1 
p q p q

m 1   
2 1 2
,

 1     1    1     1       1 1    1   
p q p q

 1 2 1 2 
21
1  2


1     1    1    1    1   
 1  2 
p q p q

m 2   
1 2 1
,

 1     1    1     1       1 1    1   
p q p q

 2 1 2 1 
12
1  1

 
1     1    1    1    1   
 1 1 
p q p q

z1   1     1  1  
2 1 2
,

1    1    1    1      1 1    1   
p q p q
  
  1 2 1 2 
21
1  2

 
1     1    1    1    1   
 1  2 
p q p q

z2   1     1  1  
1 2 1
,

1    1    1    1      1 1    1   
p q p q
  
  2 1 2 1 

12
1  1


1     1    1    1    1   
 1 1 
p q p q

v 1   
2 1 2
,

 1     1    1     1      1 1    1   
p q p q

 1 2 1 2 
21
1  2


1     1    1    1    1   
 1 2 
p q p q

v 2   
1 2 1
,

 1     1    1     1       1 1    1   
p q p q

 2 1 2 1 
12
1  1

 
1     1    1    1    1   
 1 1 
p q p q

u1   1     1  1  
2 1 2
,

1    1    1    1      1 1    1   
p q p q
  
  1 2 1 2 
21
1  2

 
  
1     1 2  1     1 1  1  2  1  1   
p q p q
  
u 2   1     1  1  
,
  
1     1 2  1    1 1    1 1  2  1  1  
p q p q
  
  

14
Mathematical Problems in Engineering.

Then,
2 i j 12 21

i ,j  1 1  i
(ai p  aj q ) 
1  1
(a1p  a2q ) 
1  2
(a2 p  a1q )
i j

  i j i j

     
2

  
2
       1  i      1  j   2  1  i p  j q       1  i     1  j   i p  j q   
p q p q


1  i 1  i
  i ,j  1 i ,j  1 
  i j i j 
 s 2   
,  i T (ai ), j T (aj )  i j i j ,

   
  i j  p (ai ) q (aj )
  2 
    
2
       1  i      1  j    1  i  j    1      1  i     1  j   i p  j q   
 1  p q p q
i ,j  1   2 p q 1  i 1  i

i
i j

  ii ,jj1 i ,j  1 
  i j 
 i j

1    1   1    
2
  1  j  1  i  1   
p q

p q 1  i

 i ,j  1
i j

 i j 
i I (ai ),j I (aj )  i j i j ,
 2
  1      1     1     
2
  1     1 i  1  j  ( 2  1) 1  i  1        1  1 i  1  j  1  i  1   
p q p q 1  i p q p q 1  i
j j 
 ii ,jj1 i ,j  1 
 i j 
 i j

1    1   1    
2
  1 j  1  i  1    
p q

p q 1  i

 i ,j  1
i j

 i j 
i F (ai ),j F (aj )  i j i j  .
 2
  1      1     1     
2
  1     1 i  1 j  ( 2  1) 1  i  1        1  1 i  1 j  1  i  1   
p q p q 1  i p q p q 1  i
j j 
 ii ,jj1 , 1
i j 
 i j 

We can achieve Eq. (2) is right when n  k .

(2) Supposing n  k , the equation below is right, then

k i j

i ,j  1 1  i
(ai p  aj q )
i j

  i j i j

     
k k
       1  i  
     1  j   2  1  i p  j q       1  i 
    1  j   i p  j q   
p q p q


1  i 1  i
  i ,j  1 i ,j  1 
  i j i j 
 s k  ,  i T (ai ), j T (aj )  i j i j ,

   
 1  i  (ai ) (aj ) 
i j p q
 k 
 
k
       1  i  
     1  j    1  i  j    1       1  i 
     1  j   i p  j q   
p q p q
2 p q 1  i 1  i

i ,j  1
i j

  i ,j 1 i ,j  1 
 i j i j 
 i j

1    1   
k
 1    1  j   1  i  1   
p q

p q 1  i

 i ,j  1
i j

 i j 
i I (ai ),j I (aj )  i j i j ,
 k
    
k
  1     1 i  1    1  j   ( 2  1) 1  i  1        1  1    1  i  1    1  j   1  i  1   
p q p q 1  i p q p q 1  i
j j 
 i ,j 1 i ,j  1 
i j i j 
 i j

1    1   
k

1    1 j   1  i  1    
p q

p q 1  i

 i ,j  1
i j

 i j 
i F (ai ),j F (aj )  i j i j  .
 k
    1    
k
  1     1 i  1    1 j   ( 2  1) 1  i  1        1  1 i  1    1  j   1  i  1   
p q p q 1  i p q p q 1  i
j j 
 ii ,jj1 i ,j  1 
 i j 

If n  k  1 , we need to calculate the equation below.

k 1    k 1j
k k
 k

 (ai p  aj q )   i j (ai p  aj q )   i k  1 (ai p  ak 1q )   (ak 1p  aj q )

i j

i ,j  1 1   i i ,j  1 1  i i 1 1  i j 1  k
1 1
i j i j

The mathematical induction on k are used to testify the equations below.

k
i k  1
  ak  1q ) 
p
(a sk ,
i 1  i i
1  
 i k  1 
 p (ai ) q (ak  1 )
i 1 
1  i 

  i k  1 i k  1

     
k k
       1  i  
    1  k  1   2  1  i p  k  1q 
      1  i     1  k  1   i p  k  1q   
p q p q

1  i 1  i

  i 1 i 1 
  i k  1 i k  1 ,

   
 i T (ai ), k  1 T (ak  1 )
  k 
 
k
       1  i  
    1  k  1   2  1  i p  k  1q 
    1       1  i      1  k  1   i p  k  1q   
p q 1  i p q 1  i
 

  i 1 i 1

 i k  1

 
k
   1     1 i  1    1  k  1   1  i  1   k 
p q q 1  i
p

 i 1
1

i I (ai ), k  1 I (ak  1 )  i k  1 i k  1 ,
 k
    
k
  1     1 i  1    1  k  1   ( 2  1) 1  i  1   k     1  1    1  i  1    1  k  1   1  i  1   k 
p q p q 1  i p q p q 1  i
1 1 
 i 1 i 1 
 i k  1

 
k
   1     1 i  1    1 k  1   1  i  1  k  
p q p q 1  i

 i 1
1

i F (ai ),k  1 F (ak  1 )  i k  1 i k  1 
 k
    
k
  1     1 i  1    1 k  1   ( 2  1) 1  i  1  k      1  1     1 i  1    1  k  1   1  i  1  k 
p q p q 1  i p q p q 1  i
1 1 
 i 1 i 1 

15
Mathematical Problems in Engineering.

k k  1j

j 1  k
1 1
(ak  1p  aj q )  sk   
  1 k 1k j1  p (ak 1 ) q (aj )
,
j 1  

  k  1j k  1j

     
k k
       1  k  1  
    1  j   2  1  k  1p  j q        1  k  1     1  j   k  1p  j q   
p q p q


1  k  1 1  k  1
  j 1 j 1 
  k  1 T (ak  1 ), j T (aj )  k  1j k  1j ,

   
  k 
 
k
       1  k  1  
    1  j   2  1  k  1p  j q      1       1  k  1     1  j   k  1p  j q   
p q 1  k  1 p q 1  k  1
 

  j 1 j 1 
 k  1j

 
k
   1     1  k  1  1    1  j   1   k  1  1   
p q p q 1  k  1

 
j
j 1
 k  1 I (ak  1 ),j I (aj )  k  1j k  1j ,


 k
   
k
  1     1  k  1  1    1  j   ( 2  1) 1   k  1  1        1  1     1  k  1  1    1  j   1   k  1  1   
p q p q 1  k  1 p q p q 1  k  1
j j 
 j 1 j 1 
 k  1j

 
k
   1     1 k  1  1    1 j   1  k  1  1    
p q p q 1  k  1

 j 1
j

k  1 F (ak  1 ),j F (aj )  k  1j k  1j 


 k
   
k
  1     1 k  1  1    1 j   ( 2  1) 1  k  1  1        1  1     1 k  1  1    1 j   1  k  1  1   
p q p q 1  k  1 p q p q 1  k  1
j j  
 j 1 j 1 

Therefore,
k 1 i j k i j k
i k  1 k k  1j
 (ai p  aj q )   (ai p  aj q )    ak  1q )   (ak  1p  aj q )
p
(a
i ,j  1 1  i i ,j  1 1  i i 1  i i
1 1
j 1  k 1
i j i j

  i j i j

   
k 1 k 1
 
 
     1  i     1  j   2  1  i p  j q i  i      1  i   
     1  j   i p  j q   
p q p q
1 1  i

  i ,j  1 ,j  1 
  i j i j 
 s k 1   ,  i T (ai ), j T (aj )  i j i j ,

   
 1 i ji  p (ai ) q (aj )   k 1 k 1 
       1  i      1  j    1  i  j  
   1      1  i  
    1  j   i p  j q   
p q p q
2 p q 1  i 1  i

i ,j  1
i j

  ii ,jj1 i ,j  1 
  i j 
 i j

1    1   
k 1
 1    1  j   1  i  1   
p q

p q 1  i

 i ,j  1
i j

 i j 
i I (ai ),j I (aj )  i j i j ,
 k 1
    
k 1
  1     1 i  1    1  j   ( 2  1) 1  i  1        1  1    1  i  1    1  j   1  i  1   
p q p q 1  i p q p q 1  i
j j 
 ii ,jj1 , 1
i j 
 i j 
 i j

1    1   
k 1

1    1 j   1  i  1    
p q

p q 1  i

 i ,j  1
i j

 i j 
i F (ai ),j F (aj )  i j i j  .
 k 1
    
k 1
  1     1 i  1    1 j   ( 2  1) 1  i  1        1  1    1 i  1    1  j   1  i  1   
p q p q 1  i p q p q 1  i
j j 
 ii ,jj1 , 1
i j 
 i j 

That is, If n  k  1, Eq. (2) is right. Therefore, for all n , Eq. (2) is right.
Then, Eq. (1) is right.

In the following, the properties of MVNLNWBMH operator will be proved.

1 1 1
(1) Reducibility. Let    , , , . Then
n n n
MVNLNWBMH (a1 ,a2 , an )  MVNLBMH (a1 ,a2 , an ).
1 1 1
Proof. Since    , , ,  , then according to the operations in Defin ition 10, the
n n n
result below can be obtained.

16
Mathematical Problems in Engineering.

1
 n  p q
 i j q 
MVNLNWBMH (a1 ,a2 , an )   (a  aj )
p
 i ,j 1 1   i 
i
 i j 
1
 1  p q
 n 2 
   n (ai p  aj q )
 i ,j 1 1  1 
 i j 
 n 
1
 n  p q
 1 q 
  (a  aj )
p
 i ,j 1 n(n  1) i 
 i j 
1
 n  p q
1
  q 
 n(n  1) i
( a p
 a )
,j  1
i j 
 i j 
 MVNLBMH (a1 ,a2 , an ).

(2) Idempotency. Let ai  a i  1,2, ,n  .Then MVNLNWBMH (a1 ,a2 , an )  a.

Proof. For each i, owing to ai  a , the formu la belo w is obtained on the basis of Eq. (5) in
Theorem 1.
1
 n  p q
 i j q 
MVNLNWBMH (a1 ,a2 , an )   (a  aj )
p
 i ,j 1 1   i 
i
 i j 
1
 n  p q

   i j (a p  a q )
 i ,j 1 1   
i
 i j 
1
 n
i j p  q 
p q

  a
 i ,j 1 1   
i
 i j 
1
 n  p q

 a i j 
 i ,j 1 1   
i
 i j 
 a.

(3) Commutativity. Let a ,a ,

1 2
an  be any permutation of a1 ,a2 , an  .Then
MVNLNWBMH a1 ,a2 , an   MVNLNWBMH (a1 ,a2 , an ).
Proof. Owing to a ,a ,
1 2
an  is permutation of (a1 ,a2 , an ) , then the equation below can be

obtained.

17
Mathematical Problems in Engineering.

1
 n  p q
 i j q 
MVNLNWBMH (a1 ,a2 , an )   (a  aj )
p
 i ,j 1 1   i 
i
 i j 
1
 n  p q

   i j (ai p  aj q )
 i ,j 1 1   
i
 i j 
(4) Monotonicity. Suppose
ai  s(ai ),T(ai ),I(ai ),F(ai ) (i  1,2, ,n ) and

bi  s(bi ),T(bi ),I(bi ),F(bi ) (i  1,2, ,n ) are two sets of MVNLNs,

when s(ai )  s(bi ), T (ai )  T (bi ), I(ai )  I(bi ) and F (ai )  F (bi ) for each i, then
MVNLNWBMH (a1 ,a2 , an )  MVNLNWBMH (b1 ,b2 , bn ).
Proof. (I) Linguistic term part
Owing to p ,q  0, and s(a )  s(b ) for each i, the result below is gained.
i i

 (ai )   (bi )
p p
 (ai )   (bi )
q q
and
  p(ai ) q(ai )   p(bi ) q(bi )
i j p 
  (ai ) q(ai )  i j  p(bi ) q(bi )
1  i 1  i
n i j p n i j p
 
i j
, 1 1  
 (ai ) q(ai )  
i j , 1 1  
 (bi ) q(bi )
i i
i j i j
1 1
 n   p q  n   p q

  i j
 (ai ) (ai )
p q  
  i j
 (bi ) (bi )
p q
 i ,j 1 1     i ,j  1 1   
i i
 i j   i j 
(II) True, indeterminacy and falsity membership parts
Owing to T (ai )  T (bi ), I(ai )  I(bi ) and F (ai )  F (bi ) for each i, p ,q  0, then the
following results can be proved easily.
1 1


 x ai  y ai  p q



 x bi  y bi  p q

;
1 1 1 1

 x ai   2  1 y ai   p q
    1 x a  y a  i i
 p q
 
x bi   2  1 y bi   p q
    1 x b  y b  i i
 p q

1 1 1 1

g ai 
  2  1 ha  
i
p q

 g a  ha
i i
 p q


g bi 
  2  1 hb   i
p q

 g b  hb
i i
 p q

;
1 1 1 1

g ai 
   1 ha 2
  i
p q
   1 g a  ha  i i
 p q
g bi 
   1 hb 2
 
i
p q
   1 g b  hb i i
 p q

1 1 1 1

v ai 
  2  1 ua  
i
p q

 v a  ua
i i
 p q


v bi 
  2  1 ub   i
p q

 v b  ub
i i
 p q

.
1 1 1 1

v ai 
   1 ua 2
  i
p q
   1 v a  ua  i i
 p q
v bi 
   1 ub 2
 
i
p q
   1 v b  ub i i
 p q

Where the corresponding x ,y ,g ,h ,u and v are defined in Definition 10.

(III) Comparing MVNLNWBMH (a1 ,a2 , an )with MVNLNWBMH (b1 ,b2 , bn )

Suppose a  s(a ),T (a ),I(a ),F(a )  MVNLNWBMH (a1 ,a2 , an ) and

18
Mathematical Problems in Engineering.

b  s(b ),T(b ),I(b ),F(b )  MVNLNWBMH (b1 ,b2 , bn ).

Because s(a )  s(b ),T (a )  T (b ),I(a )  I(b )and F(a )  F(b ),thus a  b .
Then MVNLNWBMH (a1 ,a2 , an )  MVNLNWBMH (b1 ,b2 , bn ).

In the following, a few special examples for MVNLNWBMH operator regarding

different values  , p and q will be explored.
(1) If q  0, then the MVNLNWBMH operator defined by equation (1) will be reduced
to the generalized multi-valued neutrosophic linguistic Hamacher weighted average
(GMVNLHWA) operator shown as below.
1
 n  p q
 i j q 
GMVNLHWA(a1 ,a2 , an )   (a  aj )
p
 i ,j 1 1   i 
i
 i j 
1
 n p
 i j p 
  a
 i ,j 1 1   i 
i
 i j 
1
 n i 1  i  p p
  ai 
 i 1 1  i 
 
1
 n p
  i ai p 
 i 1 

 s 1 ,
 n p

 i  p (ai )  

 i 1 

  1 
    x   y  p 
  i T (ai ), j T (aj )  ,
 1

   
1
   p     1  x   y   p 
  x   2
 1 y 

 1

   
1
 g    2  1 h  p   g   h  p 
i I (ai ), j I (aj )  1 ,
   
1
   p     1  g   h   p 
g   2
 1 h
 
 1

   
1

 v    2
 1 u  p
 v   u   p  
i F (ai ),j F (aj )  

1

   
1
   p     1 v   u   p
 v   2
 1 u  

 
i

 
n
1    1 1   i  
p
Where x   
i 1
  2  1 i p ,

 
n i

    1  i 
p
y  
i 1
 i p ,

h     1     1   
n i
 1  i 
p p
i
,
i 1

19
Mathematical Problems in Engineering.

  1   
i

 
n
1     2  1 1  i 
p

p
g  i
,
i 1

u     1     1   
n i
 1  i 
p p
i
,
i 1

v     1     1   
i

 
n
  2  1 1  i 
p p
i
i 1

(2) When p  1,q  0, then the operator of MVNLNWBMH defined by equation (1)
will be reduced to the multi-valued neutrosophic linguistic Hamacher weighted
arithmetic average (MVNLHWAA) operator shown as below.
n
MVNLHWAA(a1 ,a2 , an )  
i
i ai
1

 sn ,
 i(ai )
i 1

  n n


1     1  i 
  1   i  i
i 
 

 

  n
i 1 i 1
,
  i T (ai ) n
1     1  i 
    1  1   i  i
i
 

 
 i  1 
 i 1

 n

   i i 
 i 1 
i I (ai )  n n ,
    1 i i        1 i  
i

 
i 1


i 1

 n

   i i 
 i 1 
 

i F (ai ) n n
    1 i i        1 i  
i

 
i 1


i 1
 
If   1, the MVNLHWAA operator will be reduced to MVNLWAA operator shown
as below.

MVNLWAA(a1 ,a2 , an )  s n ,
 i(ai )
i 1

  n

1   i 
i
  i T (ai ) 1   ,
  i1 
  
n

i I (ai )  i i  ,
 i 1 
 n  
i F (ai )  i i  
 i 1 
If   2, the MVNLHWAA operator will be reduced to the multi-valued neutrosophic
linguistic Einstein we ighted arithmetic average (MVNLEWAA) operator shown in the
following.

20
Mathematical Problems in Engineering.

MVNLWAA(a1 ,a2 , an )  sn ,
 i(ai )
i 1

  n n

  1   i   1   i 
i i

 i 1
 

  n
i 1
,
  i T (ai ) n
 1   i  i  1   i  
 i

  
i 1

i 1

 n

 2 i i 
 i 1 
i I (ai )  n n ,
  2  i  i  


   i  i

i 1 i 1

 n

 2 i i 
 i 1 
 n 

i F (ai ) n
  2  i  i 


   i 
 
i

i 1 i 1

(3) If p  0,q  0, then the MVNLNWBMH operator defined by equation (1) will be
reduced to the multi-valued neutrosophic linguistic Hamacher weighted geometric
average (MVNLHWGA) operator shown as below.
n


i
MVNLHWGA(a1 ,a2 , an )  ai
i 1

 s n ,
  i (ai )
i 1

  n

     i i 
  i 1 
 ,
  i T (ai ) n n
    1  i i        1  i  
i

   

 i 1 i 1

 n n

1     1 i 
  1  i  i
i 



i 1 i 1


i I (ai )  n n ,
1     1 i 
    1  1  i  i
i
 


 i 1 i 1

 n n

    
1  i  i
i 



i 1
1    1  i
 
i 1


 n 

i F (ai ) n
1     1 i 
    1  1  i  i
i
 

 i 1 i 1
 
If   1, the MVNLHWGA operator will be reduced to MVNLWGA operator shown
as below.

21
Mathematical Problems in Engineering.

MVNLHWGA(a1 ,a2 , an )  s n ,
  i (ai )
i 1

  n
 
  i T (ai )   i i  ,
  i 1 
 n

1  i 
i
i I (ai ) 1   ,
 i1 
 n

1  i 
i
i F (ai ) 1    
 i1 
If   2, the MVNLHWGA operator will be reduced to the multi-valued neutrosophic
linguistic Einstein weighted geometr ic average (MVNLEWGA) operator shown as
below.

MVNLHWGA(a1 ,a2 , an )  s n ,
  i (ai )
i 1

  n

  2  i i 
  i 1 
 n ,
  i T (ai ) n
 2   i  i  


  
  i i

 i 1 i 1

 n n

  1  i   1  i 
i i

 i 1

i 1


i I (ai )  n n ,
 1  i  i  1  i  
 i

 
i 1

i 1

 n n

  1  i   1  i 
i i

 i 1

i 1


 n 

i F (ai ) n
 1  i  i  1  i 
 i

 
i 1

i1
 

(4) If   1, then the MVNLNWBMH operator defined by equation (1) will be reduced
to the MVNLNWBM operator shown as below.

22
Mathematical Problems in Engineering.

MVNLNWBM (a1 ,a2 , an )  s 1 ,

  p q
 n  i j p 
  
 1
 (ai ) q (aj ) 

 i ,j  1  i 
i j 

  1

  i j  p q 
 ,
 1   i p  j q 
n
 
 1 
1  i
  i T (ai ), j T (aj )  
 i ,j  1
 
 
i j
 
 1

  i j  p q 
 1  1  i  1   j    ,
n
 p q 1
1   1   
i
i I (ai ), j I (aj )
 
 
, 1
i j
 i j
 
 1

  i j  p q  
 1  1  i  1  j    
n
1  1  p q 1
   
i
i F (ai ),j F (aj ) 
 
i j
i j
, 1
  
 
(5) If   2, then the M VNLNW BM H operato r d efined by equat ion (1) will be reduced to th e
M VNLNW BM E, that is the s imp lificat io n o f mu lt i-v alued neut rosoph ic lingu ist ic NW BM
Einstein.

MVNLNWBMH (a1 ,a2 , an )  s 1 ,

  p q
 n  i j p 
  

 i ,j  1  1  i
 (ai ) q (aj ) 

 
i j 

  1

  2  x  y  p q 
  i T (ai ), j T (aj )  1 1 ,
   x  3y  p  q   x  y  p  q 
  
 1 1

  g  3h  p  q   g  h  p  q 
i I (ai ), j I (aj )  1 1 ,
  g  3h  p  q   g  h  p  q 
 
 1 1

 v  3u  p  q  v  u  p  q 
i F (ai ),j F (aj )  1 1 
 v  3u  p  q  v  u  p  q 
 
i j

 2   i  2   j  
n
p q q 1  i
Where x   3 i  j p
,
i ,j  1
i j

i j

 2   i  2   j  
n
p q q 1  i
y   i  j p
,
i ,j  1
i j

i j

  1  i  1  j  
n
 1  i  1   
p q p q 1
h  j
i
,
i ,j  1
i j

23
Mathematical Problems in Engineering.

i j

  1  i  1   j  
n
 3 1  i  1   
p q p q 1
g  j
i
,
i ,j  1
i j

i j

 
n
1  i  1  j   1  i  1  j 
p q p q 1
u  i
,
i ,j  1
i j

i j

  1  i  1  j  
n
 3 1  i  1   
p q p q 1
v  j
i
.
i ,j  1
i j

From the above analysis, we can obtain the MVNLNWBMH operator is more generalized.
5 The multiple criteria decision making approach based on the MVNLNWB MH operator
The proposed MVNLNW BMH operator is presented to cope with MCDM problem under mu lti-
valued neutrosophic linguistic environment in this subsection.
Suppose that A  A1 ,A2 , ,Am  represent m alternatives, and C  C 1 ,C 2 , ,C n  represent n

criteria. Let   1 ,2 , ,n  be the corresponding weights of criteria,

n
where j  0(j  1,2, ,n ), and  j
1
j
 1 .The evaluation value of the criteria

C j (j  1,2, ,n ) regarding the alternative Ai (i  1,2, ,m ) is provided by experts. Each value

is represented by MVNLNNs. Suppose that R  aij  m  n is the mu lti-valued neutrosophic linguistic


decision matrix, aij  s(aij ), T (aij ),I(aij ),F (aij )  is the evaluation information wh ich

represents the assessment value of alternative Ai(i  1,2, ,m ) on criteria C j (j  1,2, ,n )with

respect to the linguistic value s(aij ),where T (aij ) indicates the satisfaction degree, I(aij ) indicates
the indeterminacy degree and F (aij ) indicates the dissatisfaction degree.
Then, the main method for ranking and selecting the best alternative is presented in the following.
Step1. The decision matrix is normalized.
Generally, criteria in MCDM problems consist of two types: maximu m type and minimu m type,
the min imu m type should be transformed into the maximu m type for eliminat ing the influence of
distinguished types. Suppose that R  aij  m  n is the original decision matrix, which can be

normalized as follows:
 aij , for maximizing criteria

bij  
 s
  (aij )  
, T (aij ),I(aij ),F(aij ) ,for maximizing criteria

Thus, the normalized matrix B  bij  m  n is gained.

Step2. The comprehensive value of each alternative is calculated.

The comprehensive value represented by ai(i  1,2, ,m )can be obtained by utilizing the
MVNLNWBM H operator in Definit ion 10, which can aggregate the overall value fo r each alternative
with respect to all criteria.
Step3. The compared values of three functions are calculated.
According to the equations given in Defin ition 8, the score value denoted by E(ai ), the accuracy
value denoted by H(ai ) and the certainty value denoted by C(ai )can be obtained.

24
Mathematical Problems in Engineering.

Step4. The alternatives are selected.

Based on Defin ition 9, all alternatives Ai(i  1,2, ,m ) can be ranked on the basis of
E(ai ), H(ai ) and C(ai ), and the best alternative(s) can be selected.

6 An numberical example
In order to validate the effect iveness and practical of the novel approach, an investment project is
adapted from Ye.
An investment company wants to expand its business. Four alternatives will be chosen,
A1 represents auto corporation, A2 represents food corporation, A3 represents computer company
corporation), A4 represents weapon corporation. Each alternative is evaluated under three criteria,
C 1 denotes risk, C 2 denotes growth, C 3 denotes the impact of environ ment, where C 3 is the minimizing
criteria. The corresponding weighted vector is   0.35,0.25,0.4 . In real situation, the decision
maker may hesitant and give several possible value for the satisfaction, indeterminacy and
dissatisfaction regarding the alternative Ai corresponding to the criteria C j under the linguistic term
set S . Therefore, the assessment value is given in the form of M VNLNs, and the linguistic term set is
emp loyed as
S  s1 ,s2 ,s3 ,s 4 ,s5 ,s6 ,s7   extremely poor,very poor,poor,medium,good,very good,extremely good .
R  aij  4  3
The multi-valued neutrosophic linguistic decision matrix is shown as follows.
R  aij  4  3
 s ,0.3,0.4,0.5 ,0.1 ,0.3,0.4  s 6 ,0.5,0.6 ,0.2,0.3 ,0.3,0.4 s5 ,0.2,0.3 ,0.1,0.2 ,0.5,0.6 
 5 
 
 s 6 ,0.6,0.7 ,0.1,0.2 ,0.2,0.3 s5 ,0.6,0.7 ,0.1 ,0.3 s 5 ,0.6,0.7 ,0.1,0.2 ,0.1,0.2 
  
 s 6 ,0.5,0.6 ,0.4 ,0.2,0.3 s5 ,0.6 ,0.3 ,0.4 s 4 ,0.5,0.6 ,0.1 ,0.3 
 
 s 4 ,0.7,0.8 ,0.1 ,0.1,0.2  s 4 ,0.6,0.7 ,0.1 ,0.2 s 6 ,0.3,0.5 ,0.2 ,0.1,0.2,0.3 
 

6.1 The procedure using the proposed aggregation operator

Step1. The decision matrix is normalized.
Because C 3 is the min imizing criteria, wh ich should be converted to the maximizing criteria, then
the normalized decision matrix B  bij  m  n .can be obtained as follows:
B  bij  4  3


 s , 0.3,0.4,0.5 ,0.1 ,0.3,0.4
 5
 s 6 ,0.5,0.6 ,0.2,0.3 ,0.3,0.4 s 2 ,0.2,0.3 ,0.1,0.2 ,0.5,0.6  

 

 s 6 , 0.6,0.7 ,0.1,0.2 ,0.2,0.3  s5 ,0.6,0.7 ,0.1 ,0.3 s 2 ,0.6,0.7 ,0.1,0.2 ,0.1,0.2  
  

 s 6 , 0.5,0.6 ,0.4 ,0.2,0.3  s5 ,0.6 ,0.3 ,0.4 s 3 ,0.5,0.6 ,0.1 ,0.3 
 

 s 4 , 0.7,0.8 ,0.1 ,0.1,0.2
  s 4 ,0.6,0.7 ,0.1 ,0.2  s1 ,0.3,0.5 ,0.2 ,0.1,0.2,0.3 


Step2. The comprehensive value of each alternative is calculated.

Derive the comprehensive value ai(i  1,2, ,m ) of each alternative Ai(i  1,2, ,m )

by using the MVNLNWBM H operator presented in Definition 10. Here let p  q  1,  1. The
MVNLNWBMH operator is shown as below:

25
Mathematical Problems in Engineering.

MVNLNWBM (a1 ,a2 , an )  s 1 ,

 2
 n  i j 
  
 1
(ai )(aj ) 

 i ,j  1  i 
i j 

  1

  i j 2 
 ,
n

  i T (ai ), j T (aj )

 1   1   i  j  1  i
 
 i ,j  1
 
 
i j
 
 1

  i j 2 
 ,
 1  1  i  1   j  
n

1   1 
1  i
i I (ai ), j I (aj )
  
 
i ,j  1
i j  
 
1 
 
  i j 2  
 
 1  1  i  1  j  
n
1  1  1  i
i F (ai ),j F (aj )     
 
i j
i j
, 1
  
 
And we have
a1  s 3.9529 ,
0.3088,0.3515,0.3309,0.3752,0.3429,0.3869,0.3671,0.4118,0.3746,0.4199,0.4009,0.4463 ,
0.1255,0.1645,0.1482,0.19 ,
0.3768,0.4171,0.4049,0.446,0.4109,0.4511,0.4382,0.4783 ;

a2  s 3.9904 ,
0.6,0.6368,0.6275,0.6648,0.6345,0.6727,0.6622,0.7 ,
0.1,0.1363,0.1335,0.1714 ,
0.1882,0.2268,0.2261,0.2629 ;

a3  s 4.4850 ,
0.5272,0.5644,0.5618,0.6 ,
0.263 ,
0.2918,0.3273 ;

Step3. The compared values of three functions are calculated.

By using equations in Defin ition 8, we can obtain E(ai ) , H(ai ) and C(ai ),(i  1,2,3,4) as
follows:
E(a1 )  s2.3605 ,H (a1 )  s 0.2036 ,C(a1 )  s1.4879;
E(a2 )  s3.0440 ,H (a2 )  s1.6912 ,C(a2 )  s 2.5930;
E(a3 )  s2.9763 ,H (a3 )  s1.1382 ,C(a3 )  s 2.5267;
E(a4 )  s2.0002 ,H (a4 )  s1.0641 ,C(a4 )  s1.5461;
Step4. The alternatives are selected.
By using the compared approach in Definition 9, E(a2 ) E(a3 ) E(a1 ) E(a4 ) can

26
Mathematical Problems in Engineering.

be obtained, so the final raking of alternatives is A2 A3 A1 A4 . Apparently, A2 is the

best one, and A4 is the worst one. We don’t need to co mpare the other functions because the values of
score function differ.
6.2 Comparison analysis
We t ake d ifferen t value into cons iderat ion in step 2 to select the alternat ive fo r d iscussing th e
impact of different values p ,q and  . The comparisons are presented in Table 1 and Table 2.

Table 1 Ranking of alternatives utilizing different p ,q and   1

  1,p ,q Score function E(ai ),(i  1,2,3,4) Ranking
p  0,q  0 E(a1 )  s2.1214 ,E(a2 )  s2.8131 ,E(a3 )  s 2.8732 ,E(a4 )  s1.7094 A3 A2 A1 A4
p  0.01,q  0 E(a1 )  s2.1907 ,E(a2 )  s2.8460 ,E(a3 )  s 2.9575 ,E(a4 )  s1.7629 A3 A2 A1 A4
p  0.1,q  0 E(a1 )  s2.2165 ,E(a2 )  s2.8790 ,E(a3 )  s 2.9713 ,E(a4 )  s1.8005 A3 A2 A1 A4
p  1,q  0 E(a1 )  s2.4697 ,E(a2 )  s3.1970 ,E(a3 )  s3.1084 ,E(a4 )  s2.1573 A2 A3 A1 A4
p  2,q  0 E(a1 )  s2.7154 ,E(a2 )  s3.4882 ,E(a3 )  s3.2507 ,E(a4 )  s 2.4519 A2 A3 A1 A4
p  5,q  0 E(a1 )  s3.1768 ,E(a2 )  s3.9655 ,E(a3 )  s3.5749 ,E(a4 )  s 2.8445 A2 A3 A1 A4
p  10,q  0 E(a1 )  s3.5203 ,E(a2 )  s 4.2619 ,E(a3 )  s3.8673 ,E(a4 )  s3.0415 A2 A3 A1 A4
p  0,q  1 E(a1 )  s2.6253 ,E(a2 )  s3.3082 ,E(a3 )  s3.1563 ,E(a4 )  s 2.2904 A3 A2 A1 A4
p  0.01,q  1 E(a1 )  s2.6151 ,E(a2 )  s3.2980 ,E(a3 )  s3.1492 ,E(a4 )  s 2.2778 A2 A3 A1 A4
p  0.1,q  1 E(a1 )  s2.5377 ,E(a2 )  s3.2202 ,E(a3 )  s3.0964 ,E(a4 )  s 2.1839 A2 A3 A1 A4
p  1,q  1 E(a1 )  s2.3605 ,E(a2 )  s3.0440 ,E(a3 )  s 2.9763 ,E(a4 )  s 2.0002 A2 A3 A1 A4
p  2,q  1 E(a1 )  s2.4824 ,E(a2 )  s3.1918 ,E(a3 )  s3.0464 ,E(a4 )  s 2.1664 A2 A3 A1 A4
p  5,q  1 E(a1 )  s2.9208 ,E(a2 )  s3.6644 ,E(a3 )  s3.3532 ,E(a4 )  s 2.5870 A2 A3 A1 A4
p  10,q  1 E(a1 )  s3.3239 ,E(a2 )  s 4.0427 ,E(a3 )  s3.6946 ,E(a4 )  s 2.8680 A2 A3 A1 A4
p  0,q  2 E(a1 )  s2.8597 ,E(a2 )  s3.5712 ,E(a3 )  s3.2871 ,E(a4 )  s 2.5589 A2 A3 A1 A4
p  0.01,q  2 E(a1 )  s2.8519 ,E(a2 )  s3.5630 ,E(a3 )  s3.2820 ,E(a4 )  s2.5494 A2 A3 A1 A4
p  0.1,q  2 E(a1 )  s2.7889 ,E(a2 )  s3.4959 ,E(a3 )  s3.2400 ,E(a4 )  s 2.4731 A2 A3 A1 A4
p  1,q  2 E(a1 )  s2.5228 ,E(a2 )  s3.2087 ,E(a3 )  s3.0542 ,E(a4 )  s 2.1890 A2 A3 A1 A4
p  2,q  2 E(a1 )  s2.5363 ,E(a2 )  s3.2305 ,E(a3 )  s3.0440 ,E(a4 )  s 2.2430 A2 A3 A1 A4
p  5,q  2 E(a1 )  s2.8535 ,E(a2 )  s3.5768 ,E(a3 )  s3.2558 ,E(a4 )  s 2.5581 A2 A3 A1 A4
p  10,q  2 E(a1 )  s3.2306 ,E(a2 )  s3.9375 ,E(a3 )  s3.5836 ,E(a4 )  s 2.8212 A2 A3 A1 A4

Table 2 Ranking of alternatives utilizing different p ,q and   2

  2,p ,q Score function E(ai ),(i  1,2,3,4) Ranking
p  0,q  0 E(a1 )  s2.1344 ,E(a2 )  s2.8137 ,E(a3 )  s 2.8863 ,E(a4 )  s1.7187 A3 A2 A1 A4
p  0.01,q  0 E(a1 )  s2.1944 ,E(a2 )  s2.8478 ,E(a3 )  s 2.9631 ,E(a4 )  s1.7661 A3 A2 A1 A4
p  0.1,q  0 E(a1 )  s2.2171 ,E(a2 )  s2.8803 ,E(a3 )  s 2.9752 ,E(a4 )  s1.8021 A3 A2 A1 A4
p  1,q  0 E(a1 )  s2.4553 ,E(a2 )  s3.1938 ,E(a3 )  s3.0982 ,E(a4 )  s 2.1483 A2 A3 A1 A4
p  2,q  0 E(a1 )  s2.7142 ,E(a2 )  s3.4825 ,E(a3 )  s3.2361 ,E(a4 )  s 2.4453 A2 A3 A1 A4

27
Mathematical Problems in Engineering.

p  5,q  0 E(a1 )  s3.2212 ,E(a2 )  s3.9697 ,E(a3 )  s3.5908 ,E(a4 )  s 2.8608 A2 A3 A1 A4

p  10,q  0 E(a1 )  s3.5770 ,E(a2 )  s 4.2923 ,E(a3 )  s3.9217 ,E(a4 )  s3.0737 A2 A3 A1 A4
p  0,q  1 E(a1 )  s2.6103 ,E(a2 )  s3.3048 ,E(a3 )  s3.1457 ,E(a4 )  s 2.2818 A2 A3 A1 A4
p  0.01,q  1 E(a1 )  s2.6011 ,E(a2 )  s3.2948 ,E(a3 )  s3.1394 ,E(a4 )  s 2.2700 A2 A3 A1 A4
p  0.1,q  1 E(a1 )  s2.5312 ,E(a2 )  s3.2188 ,E(a3 )  s3.0925 ,E(a4 )  s 2.1809 A2 A3 A1 A4
p  1,q  1 E(a1 )  s2.3702 ,E(a2 )  s3.0464 ,E(a3 )  s2.9849 ,E(a4 )  s 2.0055 A2 A3 A1 A4
p  2,q  1 E(a1 )  s2.4994 ,E(a2 )  s3.1940 ,E(a3 )  s3.0556 ,E(a4 )  s 2.1751 A2 A3 A1 A4
p  5,q  1 E(a1 )  s2.9637 ,E(a2 )  s3.6734 ,E(a3 )  s3.3833 ,E(a4 )  s 2.6091 A2 A3 A1 A4
p  10,q  1 E(a1 )  s3.3787 ,E(a2 )  s 4.0718 ,E(a3 )  s3.7516 ,E(a4 )  s 2.9010 A2 A3 A1 A4
p  0,q  2 E(a1 )  s2.8577 ,E(a2 )  s3.5652 ,E(a3 )  s3.2720 ,E(a4 )  s 2.5525 A2 A3 A1 A4
p  0.01,q  2 E(a1 )  s2.8506 ,E(a2 )  s3.5572 ,E(a3 )  s3.2675 ,E(a4 )  s 2.5435 A2 A3 A1 A4
p  0.1,q  2 E(a1 )  s2.7919 ,E(a2 )  s3.4917 ,E(a3 )  s3.2304 ,E(a4 )  s 2.4707 A2 A3 A1 A4
p  1,q  2 E(a1 )  s2.5398 ,E(a2 )  s3.2108 ,E(a3 )  s3.0635 ,E(a4 )  s2.1976 A2 A3 A1 A4
p  2,q  2 E(a1 )  s2.5586 ,E(a2 )  s3.2343 ,E(a3 )  s3.0581 ,E(a4 )  s 2.2559 A2 A3 A1 A4
p  5,q  2 E(a1 )  s2.8950 ,E(a2 )  s3.5871 ,E(a3 )  s3.2878 ,E(a4 )  s 2.5824 A2 A3 A1 A4
p  10,q  2 E(a1 )  s3.2843 ,E(a2 )  s3.9650 ,E(a3 )  s3.6392 ,E(a4 )  s 2.8549 A2 A3 A1 A4
In tab le 1，we t ake the p arameter value   1,which is based on Algebraic operation, and t he
MVNLNWBM operator is applied. In t ab le 2，we take th e p arameter valu e   2,which is
based on Einstein operation, and t he MVNLNW BM E operator is applied. As we can see from
table 1 and table 2, the ordering of alternatives taking different p arameters p ,q and  may
be different, because the different parameters will cause d ifferent sco re fun ct ion valu e.
Ho wever, A2 or A3 is always the best select ion , and A4 is always th e wo rst select ion. Wh ether
  1 or   2, the same ran king results is obt ained with reg ard to th e same p arameter
value p and the same paramet er value q except fo r on e s itu at ion in
wh ich p  0,q  1 and   1 .Specially , if p  0,q  0, M VNLNW BM H will reduce to
M VNLHW GA ope rato r. If p  0,q  0 and   1 , M VNLHW GA will reduce to M VNLW GA
operator. When p  1,q  0 , M VNLNW BMH will reduce to M VNLHWAA operator. If
p  1,q  0 and   1 , M VNLHWAA will reduce to MVNLWAA operator. If   1 in Table 1,
the ranking results on the basis of the M VNLW GA and M VNLWAA operators differ, wh ich due to the
two operators emphasis on different major points, and the same situation happens in Table 2.When the
parameters q  0 and p  0,p  0.01,p  0.1,respectively, the rankings are identical in t wo
Tables, the ranking order is always A3 A2 A1 A4 . When p and q are assign the other values in
two Tables, the ran king order is changed, and the result is A2 A3 A1 A4 .That is, the best
selection is from A3 to A2 except for one situation where p  0,q  1 in Table 1.
For illustrating the effective and flexib le of the novel approach, the method in literature is adopted in
mu lti-valued neutrosophic linguistic environ ment in this paper, and the same ranking orders are
obtained in ref [8] where the SVNHFWA and SVNHFW G operators are adopted to fuse single-valued
neutrosophic hesitant fuzzy information. When   1,p  0,q  0 and   1,p  0,q  0, the
two operators are special cases of M VNLNWBM H operator .Therefore, novel operator in this paper
has better flexib ility and generalizat ion. In actual cases, the decision makers can assign different

28
Mathematical Problems in Engineering.

parameter values  , p and q .. Generally, for convenience, we can set   p  1,which can not only

simplify the calculation, but also consider the interrelationship of multiple values.

7 Conclusions
In this paper, the MVNLS are proposed by combining the M VNS and LS, wh ich not only describe
linguistic terms, but also give the quantitative value of three membership degrees concerning the
linguistic variables, which has better flexibility to express the decision informat ion. Moreover, NW BM
is a useful operator which has the trait of taking into account the interrelationship of different
arguments, and overcome drawbacks of non-reducib ility and non-idempotency. Hamacher operations
are the extension of Algebraic and Einstein operations, which is more general. Considering these
advantages, we have developed Hamacher operational laws for M VNLNs and extended the NWBM to
fuse MVNL in formation. Thus, the M VNLNW BMH operator is proposed, which is appropriate to deal
with M VNL info rmation. So me desirable properties of the novel operator are discuss ed in detail, and
some special cases are analy zed. Furthermore, the co mparison method for M VNLNs is also studied,
and the rankings of alternatives affecting by d ifferent parameters p ,q and  are also compared. Fo r
verify ing the novel approach, we successfully applied the approach to an example. The results show
the novel approach has the follo wing advantages. The M VNLNWBM H operator is more flexib le and
more general, and which can capture the interrelationship among argu ments and express decision
informat ion more practical, the decision makers can assign appropriate values according to the real
situation. In future, we will exp lore to apply the operator to the different do mains, for instance, fault
diagnosis, machine learning and medical diagnosis.
As a future possible research, we will extend our research by using the refined neutrosophic set [61],
i.e. the truth value T is refined into types of sub -truths such as T1, T2, etc.,
similarly indeterminacy I is refined into types of sub-indeterminacies I1, I2, etc., and the sub-falsehood
F is split into F1, F2, etc.

Acknowledgement
This paper is supported by the National Natural Science Foundation of Ch ina (No. 71371154), the
Hu manities and Social Sciences Research Pro ject of Min istry of Education of China (No.
16YJCZH049), Th inking Bank Project of Hubei (No. HBSXK2017055) and the Hu manit ies and social
Sciences foundation of Depart ment of Education of Hubei (No. 17Q121).The authors also would like to
express appreciation to the anonymous reviewers and Editor for their helpfu l co mments that imp roved
the paper.
Compliance with ethical standards
Conflict of interest ：The authors declare that they have no conflict of interests regarding the
publication of this paper.

References：
1. Zadeh LA (1965) Fuzzy sets. Inf Control 8:338–356
2. Atanassov KT (1986) Intuitionistic fuzzy sets. Fuzzy Sets Syst 20:87–96
3. Torra V (2010) Hesitant fuzzy sets. Int J Intell Syst 25:529–5394

29
Mathematical Problems in Engineering.

4. Smarandache F (1995) Neutrosophy, Neutrosophic Set, Log ic, Probability and Statistics .
University of New Mexico, http://fs.gallup.unm.edu/neutrosophy.htm
5. Wang H, Smarandache F, Zhang YQ, Sunderraman R (2010) Single valued neutrosophic sets.
Multisp Multistruct 4:410–413
6. Ye J (2014) A mu lticriteria decision-making method using aggregation operators for simplified
neutrosophic sets. J Intell Fuzzy Syst 26:2459–2466
7. Wang H, Smarandache F, Zhang Y-Q, Sunderraman R (2005) Interval neutrosophic sets and logic:
theory and applications in computing. Hexis, Arizona
8.Ye J (2015) Multip le-attribute decision-making method under a single-valued neutrosophic hesitant
fuzzy environment. J Intell Syst 24(1):23– 36
9.Wang JQ, Li XE (2015) An application of the TODIM method with mu lt i-valued neutrosophic set.
Control Decis 30(6):1139–1142
10. Liu P, Wang Y (2014) Multip le attribute decision -making method based on single-valued
neutrosophic normalized weighted Bonferron i mean. Neu ral Co mput Appl. doi:10.1007/s00521-014-
1688-8
11. Ye J (2014) Single valued neutrosophic cross -entropy for mu lti-criteria decision making problems.
Appl Math Model 38(3):1170–1175
12. Zhang HY, Wang JQ, Chen XH (2014) Interval neutrosophic sets and their application in
multicriteria decision making problems. Sci World J 2014(1):645953.
13. Liu P, Shi L (2014) The generalized hybrid weighted average operator based on interval
neutrosophic hesitant set and its application to mult iple attribute decision making. Neural Co mput Appl.
doi:10.1007/s00521-014-1736-4
14. Brou mi S, Smarandache F (2013) Correlat ion coefficient of interval neutrosophic set. Appl Mech
Mater 436:511–517
15. Brou mi S, Smarandache F (2014) Cosine similarity measure of interval valued neutrosophic sets.
Neutrosophic Sets Syst 5:15–20
16. Brou mi S, Smarandache F (2015) New Operat ions on Interval Neutrosophic Sets. Journal of New
Theory 1:24–37
17. Ye J (2014) Similarity measures between interval neutrosophic sets and their applications in
multicriteria decision-making. J Intell Fuzzy Syst 26:165–172
18. Peng JJ, Wang JQ, Wu XH, Wang J, Chen X H (2015) Multi-valued neutrosophic sets and power
aggregation operators with their applicat ions in mult i-criteria group decision-making problems. Int J
Comput Intell Syst 8(2):345–363
19. Peng JJ, Wang JQ, Wu XH (2016) An extension of the ELECTRE approach with mult i-valued
neutrosophic information. Neural Comput Appl. doi:10.1007/s00521-016-2411-8
20. Ji P, Zhang HY, Wang JQ (2016) A pro jection-based TODIM method under mult i-valued
neutrosophic environments and its application in personnel selection. Neural Co mput Appl.
doi:10.1007/s00521-016-2436-z
21. Zadeh LA (1975) The concept of a linguistic variable and its application to appro ximate reasoning.
Inf Sci 8(3):199– 249
22. Wang JQ, Wang J, Chen QH, Zhang HY, Chen XH (2014) An outranking approach for mu lti-
criteria decision-making with hesitant fuzzy linguistic term sets. Inf Sci 280:338–351
23. Wang J, Wang JQ, Zhang HY (2016) A likelihood-based TODIM approach based on mult i-hesitant

30
Mathematical Problems in Engineering.

fuzzy linguistic information for evaluation in logistics outso urcing. Comput Ind Eng 99:287–299
24.Wang J, Wang JQ, Zhang HY, Chen XH (2016) Mu lti-criteria group decision making approach
based on 2-tuple linguistic aggregation operators with mult i-hesitant fuzzy linguistic information. Int J
Fuzzy Syst 18(1):81– 97
25. Meng FY, Chen Xh, Zhang Q (2014) Mu lti-attribute decision analysis under a linguistic hesitant
fuzzy environment. Inf Sci 267:287-305
26. Tian ZP, Wang J, Wang JQ, Chen XH (2015) Mult i-criteria decision-making approach based on
gray linguistic weighted Bonferroni mean operator. Int Trans Oper Res. doi:10.1111/ itor. 12220
27. Ye J (2015) An extended TOPSIS method for mu ltip le attribute group decision making based on
single valued neutrosophic linguistic numbers. J Intell Fuzzy Syst 28(1):247–255
28. Brou mi S, Smarandache F (2014) Single valued neutrosophic trapezoid linguistic aggregation
operators based multi-attribute decision making. Bull Pure Appl Sci Math Stat 33(2):135– 155
29. Tian ZP, Wang J, Zhang HY, Chen XH, Wang JQ (2015) Simp lified neutrosoph ic linguistic
normalized weighted Bonferroni mean operator and its application to mult i-criteria decision making
problems. Filo mat. doi:10.2298/FIL1508576F
30. Tian ZP, Wang J, Zhang HY, Wang JQ (2016) Multi-criteria decision-making based on generalized
prioritized aggregation operators under simplified neutrosophic uncertain linguistic environ ment. Int J
Mach Learn Cybern. doi:10.1007/s13042-016-0552-9
31. Ye J (2014) So me aggregation operators of interval neutrosophic linguistic nu mbers for mu ltip le
attribute decision making. J Intell Fuzzy Syst 27(5):2231–2241
32. Ma YX, Wang JQ, Wang J, Wu XH (2016) An interval neutrosophic linguistic mu lt i-criteria g roup
decision-making method and its application in selecting medical treat ment options. Neural Co mput
Appl. doi:10.1007/s00521-016-2203-1
33. Broumi S, Ye J, Smarandache F (2015) An extended TOPSIS method for mu ltip le attribute
decision making based on interval neutrosophic uncertain linguistic variables. Neutrosophic Sets Syst
8:23–32
34. Yager RR (2008) Prioritized aggregation operators. Int J Approx Reason 48(1):263– 274
35. Xu ZS (2009) Fuzzy harmonic mean operators. Int J Intell Syst 24(2):152–172
36. Ye J (2015) Mu ltiple attribute decision-making method based on the possibility degree ranking
method and ordered weighted aggregation operators of interval neutrosophic numbers. J Intell Fu zzy
Syst 28(3):1307–1317
37.Sun HX, Yang HX, Wu JZ, Yao OY (2015) Interval neutrosophic numbers Choquet integral
operator for multi-criteria decision making. J Intell Fuzzy Syst 28(6):2443– 2455
38.Liu PD, Wang YM (2015) Interval neutrosophic prioritized OWA operator and its application to
multiple attribute decision making. J Syst Sci Complex. doi:10.1007/s11424-015-4010-7
39. Liu PD, Teng F (2015) Multip le attribute decision making method based on normal neutrosophic
generalized weighted power averaging operator. Int J Mach Learn Cybernet. doi:10.1007/s13042-015-
0385-y
40. Li DF (2011) The GOWA operator based approach to multi -attribute decision making using
intuitionistic fuzzy sets. Math Comput Model 53:1182–1196
41. Wei GW (2010) So me induced geometric aggregation operators with in tuitionistic fuzzy
information and their application to group decision making. Appl Soft Comput 10:423–431
42. Bonferroni C (1950) Sulle medie multiple di potenze. Boll Mat Ital 5:267–270

31
Mathematical Problems in Engineering.

43. Beliakov G, Simon J (2013) On extending generalized Bonferroni means to Atanassov orthopairs in
decision making contexts. Fuzzy Sets Syst 211(1):84–98
44. Xu ZS, Yager RR (2011) Intuitionistic fuzzy Bonferroni Means, IEEE transactions on systems,
man, and cybernetics—part B. Cybernetics 41:568–578
45. Zhou W, He JM (2012) Intuitionistic fuzzy normalized weighted Bonferroni mean and its
application in mult icriteria decision making. J Appl Mat 2012:1–22
46. Liu PD, Jin F (2012) The t rapezoid fu zzy linguistic Bonferroni mean operators and their
application to multiple attribute decision making. Sci Iran Trans E 19(6):1947–1959
47. Liu PD, Wang YM (2014) Multip le attribute decision -making method based on single-valued
neutrosophic normalized weighted Bonferroni mean. Neural Comput Appl 25(7–8):2001–2010
48. Liu PD, Chen YB, Chu YC (2014) Intuitionistic uncertain linguistic weighted Bonferroni OWA
operator and its application to mult iple attribute decision making. Cybern Syst 45(5):418–438
49. Liu PD, Rong LL, Chu YC, Li YW (2014) Intuitionistic linguistic weighted Bonferroni mean
operator and its application to multiple attribute decision making. Sci World J 2014:1–13
50.Liu PD, Sh i LL (2015) So me neutrosophic uncertain linguistic number Heronian mean operators
and their applicat ion to multi-attribute group decision making. Neural Co mput Appl.
doi:10.1007/s00521-015-2122-6
51.Liu PD (2014) So me Hamacher aggregation operators based on the interval-valued intuitionistic
fuzzy numbers and their application to Group Decision Making. IEEE Trans Fuzzy Syst 22(1):83– 97
52. Liu PD, Chu YC, Li YW, Chen YB (2014) So me generalized neutrosophic number Hamacher
aggregation operators and their application to group decision making. Int J Fuzzy Syst 16(2):242– 255
53. Xu ZS (2006) A note on linguistic hybrid arith metic averaging operator in mu ltiple attribute group
decision making with linguistic information. Group Decis Negot 15(6):593-604
54. Smarandache F (1998) A Unifying Field in Logics.Neutrosophy: Neutrosophic Probab ility, Set
and Logic.Rehoboth: American Research Press.

32