mathematics

Article
A Study of Regular and Irregular Neutrosophic
Graphs with Real Life Applications
Liangsong Huang 1, *, Yu Hu 1 , Yuxia Li 1 , P. K. Kishore Kumar 2 , Dipak Koley 3 and
Arindam Dey 4
1 Key Laboratory for Robot Intelligent Technology of Shandong Province, Shandong University of Science and
Technology, Qingdao 266590, China; [email protected] (Y.H.); [email protected] (Y.L.)
2 Information Technology Department, Al Musanna College of Technology, Sultanate of Oman 111, Oman;
[email protected]
3 Department of Computer Application, Techno India Hooghly, Chinsurah 712101, West Bengal, India;
[email protected]
4 Department of Computer Science and Engineering, Saroj Mohan Institute of Technology, Hooghly 712512,
West Bengal, India; [email protected]
* Correspondence: [email protected]




Received: 23 April 2019; Accepted: 3 June 2019; Published: 17 June 2019


Abstract: Fuzzy graph theory is a useful and well-known tool to model and solve many real-life
optimization problems. Since real-life problems are often uncertain due to inconsistent and
indeterminate information, it is very hard for an expert to model those problems using a fuzzy
graph. A neutrosophic graph can deal with the uncertainty associated with the inconsistent and
indeterminate information of any real-world problem, where fuzzy graphs may fail to reveal
satisfactory results. The concepts of the regularity and degree of a node play a significant role
in both the theory and application of graph theory in the neutrosophic environment. In this work,
we describe the utility of the regular neutrosophic graph and bipartite neutrosophic graph to model
an assignment problem, a road transport network, and a social network. For this purpose, we
introduce the definitions of the regular neutrosophic graph, star neutrosophic graph, regular complete
neutrosophic graph, complete bipartite neutrosophic graph, and regular strong neutrosophic graph.
We define the dm - and tdm -degrees of a node in a regular neutrosophic graph. Depending on the
degree of the node, this paper classifies the regularity of a neutrosophic graph into three types, namely
dm -regular, tdm -regular, and m-highly irregular neutrosophic graphs. We present some theorems and
properties of those regular neutrosophic graphs. The concept of an m-highly irregular neutrosophic
graph on cycle and path graphs is also investigated in this paper. The definition of busy and free
nodes in a regular neutrosophic graph is presented here. We introduce the idea of the µ-complement
and h-morphism of a regular neutrosophic graph. Some properties of complement and isomorphic
regular neutrosophic graphs are presented here.

Keywords: neutrosophic graph; complete neutrosophic graph; bipartite neutrosophic graph;

complement neutrosophic graph; road transport network; wireless multihop network
and social network

PACS: J0101

1. Introduction
Graph theory has many real-life applications for problems in computer applications, systems
analysis, computer networks, transportation, operations research, and economics. A graph is basically
a model of relations, and it is a used to represent the real-life problem consisting of relationships

Mathematics 2019, 7, 551; doi:10.3390/math7060551 www.mdpi.com/journal/mathematics

Mathematics 2019, 7, 551 2 of 20

between objects. The vertices and edges of the graph are used to represent the objects and the relations
between objects, respectively. In many optimization problems, the available information is inexact or
imprecise for various reasons such as the loss of information, a lack of evidence, imperfect statistical
data and insufficient information. Generally, the uncertainty in real-life problems may be present in
the information that defines the problem.
Classical graph theory uses the basic concept of classical set theory, which was presented by
Cantor. In a classical graph, for any vertex or edge, there are two possibilities: it is either in the graph
or it is not in the graph. Therefore, classical graphs cannot model uncertain optimization problems.
Real-life problems are often uncertain, which are hard to model using classical graphs. The fuzzy set [1]
is an extended version of the classical set, where the objects have varying membership degrees. A fuzzy
set gives its objects different membership degrees between zero and one. The membership degree is
not the same as probability; rather, it describes membership in vaguely-defined sets. The concept of
fuzziness in graph theory [2] was described by Kaufmann [3] using the fuzzy relation. Rosenfeld [4]
introduced some concepts such as bridges, cycles, paths, trees, and the connectedness of the fuzzy
graph and described some of the properties of the fuzzy graph. Gani and Radha [5] presented the
notation of the regular fuzzy graph. Many other researchers, such as Samanta and Pal [6], Rashmanlou
and Pal [7], Rashmanlou et al. [8,9], Paramik [10], Nandhini [11], Ghorai and Pal [12], Dey et. al [13],
and Borzooei [14] presented much work in the domain of the fuzzy graph and its several applications
in real life. Samanta and Pal [6] and Rashmanlou and Pal [15] presented the concept of the irregular
and regular fuzzy graph. They also described some applications of those graphs.
Atanassov [16–18] presented the idea of the intuitionistic fuzzy set (IFS) as a modified version of
the classical fuzzy set. The classical fuzzy set uses only the membership grade (degree), but IFS uses
independent membership grade and non-membership grade for any entity, and the only requirement
is that the sum of non-membership and membership degree values be no greater than one. Shannon
and Atanassov [19] presented the idea of the IFS relation and the intuitionistic fuzzy graph (IFG)
and discussed many theorems, proofs, and proprieties [19]. Parvathi et al. [20–22] presented many
different operations such as the join, union, and product of two IFGs. Rashmanlou et al. [23] presented
some products such as strong, direct, and lexicographic products for two IFGs. For further detailed
descriptions of IFGs, please refer to [24–27]. Although the fuzzy graph and IFG have been used to
model many real-word problems, uncertainties due to inconsistent and indeterminate information
about a problem cannot be represented properly by the fuzzy graph or IFG. Therefore, a new idea is
needed to deal with such scenarios.
Smarandache [28] introduced the idea of the neutrosophic set, by modifying the concept of the
fuzzy set. The neutrosophic set can work with uncertain, indeterminate, vague, and inconsistent
information of any uncertain real-life problem. It is basically a modified version of the crisp set, Type
1 fuzzy set, and IFS. It is described by the truth, indeterminate, and false membership degrees of
any object. These three membership degrees are independent and always lie within ]− 0, 1+ [, i.e.,
a nonstandard unit interval. The neutrosophic graph [12] can efficiently model the inconsistent
information about any real-life problem. Recently, many researchers have more actively worked on
neutrosophic graph theory; for instance, Ye [29], Yang et al. [30], Naz et al. [31], Broumi [15], and
Arkam [32–35].
The vertex degree is a significant way to represent the total number of relations of a vertex in
a graph, and the vertex degree can be used to analyze the graph. Gani and Lathi [36] proposed the
concept of irregularity, total irregularity, and total degree in a fuzzy graph. Maheswari and Sekar [37]
proposed the notation of the d2 -vertex in a fuzzy graph and also described several properties on the
d2 -vertex degree of a fuzzy graph. Darabian et al. [38] presented the idea of the dm -regular vague graph,
tdm -regular vague graph, m-highly irregular vague graph, and m-highly total irregular vague graph,
and they discussed some properties of those graphs. They described some real-life applications (e.g.,
fullerene molecules, wireless networks, and road networks) of regular vague graphs. Neutrosophic
graphs are more effective, precise, flexible, and compatible when modeling uncertain real-life problems
Mathematics 2019, 7, 551 3 of 20

compared to fuzzy graphs or vague graphs. Thus, the use of the neutrosophic graph is inevitable for
modeling optimization problems [29,34,39–41] in real-life scenarios, and it is essential to present some
new properties and theories for neutrosophic graphs. This idea motivates us to introduce different
types of neutrosophic graphs (regular, bipartite, isomorphic, and µ complement neutrosophic graphs)
and their related theorems.
The concept of the regularity and degree of nodes has a significant role in both theories and
applications (e.g., social network analysis, road transpotation network, wireless multihop network,
and the assignment problem) in neutrosophic graph theory. The main contributions of this manuscript
are as follows.

(i) As far as we know, there exists no research work on the regularity of the neutrosophic graph
until now. Therefore, in this manuscript, we present the definition of the regular neutrosophic
graph, star neutrosophic graph, regular strong neutrosophic graph, and complete bipartite
neutrosophic graph.
(ii) We introduce the two types of degree, dm and total dm -degrees, of a node in a neutrosophic
graph. The definition of busy and free nodes in a regular neutrosophic graph are presented here.
(iii) Depending on the degree of the node, we introduce the three types of regularity of a
neutrosophic graph, dm -regular, tdm -regular, and m-highly-irregular neutrosophic graphs
(HING), and some properties are also described. We also investigate the concept of m-HING
on the cycle graph and path graph in this manuscript.
(iv) We have presented the definition of the µ-complement and h-morphism of a neutrosophic
graph. Some properties of the complement and isomorphic regular neutrosophic graph are
also presented here.
(v) Some real-life applications of the regular neutrosophic graph and complete bipartite
neutrosophic graph such as the road transport network, social networks, and assignment
problem are described in this paper.

2. Preliminaries
In this portion, we will discuss the single-valued neutrosophic graph (SVNG), adjacent node,
path, isolated node, strength of a path, strong SVNG, complement SVNG, and complete SVNG, which
are efficient for the present work.

Definition 1. Let G = ( N, M) be an SVNG [42,43] where N and M are represented by two neutrosophic sets
on V and E, respectively, which satisfy the following.

TM ( a, b) ≤ min ( TN ( a) , TN (b))
I M ( a, b) ≥ max ( IN ( a) , IN (b))
FM ( a, b) ≥ max ( FN ( a) , FN (b))

Here, a and b are two vertices of G, and ( a, b) ∈ E is an edge of G.

Definition 2. The two nodes a and b of an SVNG G = ( N, M) are defined as the adjacent nodes in G if and
only if ( TM ( a, b), I M ( a, b), FM ( a, b)) > 0. The two nodes a and b are called the neighbor nodes, and ( a, b) is
incident at a and b.

Definition 3. Let G = ( N, M) be an SVNG and P be a path of G. P is a collection of different nodes,

a0 , a1 , a2 , ..., an such that ( TM ( ai−1 , ai ) , I M ( ai−1 , ai ) , FM ( ai−1 , ai )) > 0 for 0 ≤ i ≤ n. Here, n represents
the neutrosophic length of the path P. A single neutrosophic vertex, i.e., ai in G, is also assumed as a path.
The path length of a single node ai is (0, 0, 0). We define the order pair ( ai−1 , ai ) as the edge of the path. P is said
to be a neutrosophic cycle if a0 = an and n ≥ 3.

Definition 4. A node ai ∈ V of an SVNG G is called the isolated node if there exists no incident arc to
the vertex ai .
Mathematics 2019, 7, 551 4 of 20

Definition 5. Let G = ( N, M ) be an SVNG. If G has a path P of path length [42] k from node a to node b in
k ( a, b ), I k ( a, b ), and F k ( a, b ) are
G such as P = { a = a1 , ( a1 , a2 ), a2 , ..., ak−1 ( ak−1 , ak ), ak = b)}, then TM M M
described as follows.

TM k ( a, b ) = sup ( T ( a, a ) ∧ T ( a , a ) ∧ ... ∧ T ( a
M 1 M 1 2 B k −1 , ak ))
k
I M ( a, b) = inf ( I M ( a, a1 ) ∨ I M ( a1 , a2 ) ∨ ... ∨ I M ( ak−1 , ak ))
FMk ( a, b ) = inf ( F ( a, a ) ∨ F ( a , a ) ∨ ... ∨ F ( a
M 1 M 1 2 M k −1 , ak ))

Definition 6. Let G = ( N, M ) be an SVNG. The strength of connection of a path P between two nodes a and
∞ ( a, b ), I ∞ ( a, b ), F ∞ ( a, b ) where:


b is defined by TM M M

TM∞ ( a, b ) = sup{ T k ( a, b ) | k = 1, 2, ...}

M
∞ k ( a, b ) | k = 1, 2, ...}
I M ( a, b) = inf{ I M
FM∞ ( a, b ) = inf{ F k ( a, b ) | k = 1, 2, ...}
B

Definition 7. Let G = ( N, M ) be an SVNG. G is said to be a connected SVNG [42] if there exists no isolated
node in G.

Definition 8. Let G = ( N, M) be an SVNG and a ∈ V be node of G. The degree of node a is the sum of the
truth membership values, the sum of the indeterminacy membership values, and the sum of the membership
values of falsity of all the arcs that are adjacent to node a. The degree of node [35] a is denoted by d( a) =
(d T ( a) , d I ( a) , d F ( a)) where:

dT ( a) = ∑ TM (a, b) , d I (a) = ∑ I M ( a, b) , d F ( a) = ∑ FM ( a, b)
a ∈V a ∈V a ∈V
a6=b a6=b a6=b

Here, d T ( a), d I ( a), and d F ( a) are the degree of the truth membership value, the degree of the indeterminacy
membership value, and the degree falsity membership value, respectively, of vertex a.

Definition 9. Let G = ( N, M) be an SVNG. Then, the order of G is denoted by

O( G ) = (OT ( G ) , O I ( G ) , OF ( G )) where:

OT ( G ) = ∑ TN (a) , O I (G) = ∑ IN ( a) , OF ( G ) = ∑ FN ( a)
a ∈V a ∈V a ∈V

Here, OT ( G ), O I ( G ), and OF ( G ) are the order of the membership degree of the truth value, the
indeterminacy value, and the falsity value of G, respectively.

Definition 10. Let G = ( N, M ) be an SVNG and the size graph of G be described as

S( G ) = (ST ( G ) , S I ( G ) , S F ( G )) where:

ST ( G ) = ∑ TM ( a, b) , S I ( G ) = ∑ I M ( a, b) , S F ( G ) = ∑ FM ( a, b)
a,b∈V a,b∈V a,b∈V
a6=b a6=b a6=b

Here, ST ( G ), S I ( G ), and S F ( G ) are respectively the order of the membership degree of the truth,
indeterminacy, and falsity of G.

Definition 11. Let G = ( N, M) be an SVNG. G is said to be a strong SVNG [31,35,44] if:

TM ( a, b) = min ( TN ( a) , TN (b))
I M ( a, b) = max ( IN ( a) , IN (b))
FM ( a, b) = max ( FN ( a) , FN (b)), ∀( a, b) ∈ E

Definition 12. Let G = ( N, M) be an SVNG and G be said to be a complete SVNG [31,35,44] if:
Mathematics 2019, 7, 551 5 of 20

TM ( a, b) = min ( TN ( a) , TN (b))
I M ( a, b) = max ( IN ( a) , IN (b))
FM ( a, b) = max ( FN ( a) , FN (b)), ∀ a, b ∈ V

Definition 13. Let G = ( N, M ) be an SVNG. G c = ( N c , Mc ) is the complement of an SVNG [31,35,44] if

N c = N and Mc is computed as below.

TMc ( a, b ) = min ( T ( a ) , T ( b )) − T ( a, b )
N N M
c ( a, b ) = max ( I ( a ) , I ( b )) − I ( a, b )
IM N N M
FMc ( a, b ) = max ( F ( a ) , F ( b )) − F ( a, b ), ∀ a, b ∈ V
N N M

Here, TMc ( a, b ), I c ( a, b ), and F c ( a, b ) denote the true, intermediate, and false membership degree for edge
M M
c
( a, b) of G .

3. Regular, dm -Regular, and tdm -Regular Neutrosophic Graphs

In this section, first we define the regular neutrosophic graph, regular strong neutrosophic graph,
dm -degree, and tdm -degree of nodes in a neutrosophic graph. Then, we propose the notions of dm and
tdm -regular neutrosophic graphs and prove the necessary and sufficient conditions, for which under
these conditions, dm -regular with tdm -regular neutrosophic graphs are equivalent.

Definition 14. Let G = ( N, M ) be an SVNG. G is a regular neutrosophic graph if it satisfies the

following conditions.

∑ TM (a, b) = Constant, ∑ IM (a, b) = Constant, ∑ FM (a, b) = Constant

a6=b a6=b a6=b

Definition 15. Let G = ( N, M) be an SVNG. G is a regular strong neutrosophic graph if it satisfies the
following conditions.

TM ( a, b) = min ( TN ( a) , TN (b)) and ∑ TM (a, b) = Constant

a6=b

I M ( a, b) = max ( IN ( a) , IN (b)) and ∑ IM (a, b) = Constant

a6=b

FM ( a, b) = max ( FN ( a) , FN (b)) and ∑ FM (a, b) = Constant

a6=b

Definition 16. Let G = ( N, M) be an SVNG and the dm -degree of any node a in G be denoted as dm ( a) where:
!
dm ( a) = ∑ m
TM ( a, b) , ∑ m
IM ( a, b) , ∑ m
FM ( a, b)
a 6 = b ∈V a 6 = b ∈V a 6 = b ∈V

Here, the path a, a1 , a2 , ..., am−1 , b is the shortest path between the nodes u and v, and the length of this
path is m.

Example 1. We have considered an example of an SVNG G, presented in Figure 1.

Mathematics 2019, 7, 551 6 of 20

(.3, .6, .7) (.3, .5, .6)

a b
(.2, .6, .8)

(.2 )
, .8 , .7
, .9 , .6
) (.2

c
(.4, .5, .6)

Figure 1. A neutrosophic graph G.

Then, the d2 -degree of the nodes in G is computed as follows.

d2 ( a) = (0.2 + 0.2, 0.6 + 0.8, 0.8 + 0.9) = (0.4, 1.4, 1.7)
d2 (b) = (0.2 + 0.2, 0.8 + 0.8, 0.9 + 0.9) = (0.4, 1.6, 1.8)
d2 (c) = (0.2 + 0.2, 0.6 + 0.8, 0.8 + 0.9) = (0.4, 1.6, 1.8)

Definition 17. Let G = ( N, M ) be an SVNG and a ∈ V be a node G. The total dm -degree (tdm -degree) of
node a in G is computed as follows.
! ! !!
tdm ( a) = ∑ m
TM ( a, b) + TN ( a) , ∑ m
IM ( a, b) + IN ( a) , ∑ m
FM ( a, b) + FN ( a)
a 6 = b ∈V a 6 = b ∈V a 6 = b ∈V

Here, the path a, a1 , a2 , ..., am−1 , b is the shortest path between the nodes u and v, and the length of this
path is m.

Example 2. Let us consider an example of an SVNG, shown in Figure 1. Then, the td2 -degree of the nodes in G
is as follows.
td2 ( a) = (((0.2 + 0.2) + 0.3) , ((0.6 + 0.8) + 0.6) , ((0.8 + 0.9) + 0.7)) = (0.7, 2.0, 2.4)
td2 (b) = (((0.2 + 0.2) + 0.3) , ((0.8 + 0.8) + 0.6) , ((0.9 + 0.9) + 0.6)) = (0.7, 2.1, 2.4)
td2 ( a) = (((0.2 + 0.2) + 0.4) , ((0.8 + 0.8) + 0.5) , ((0.9 + 0.9) + 0.6)) = (0.8, 2.1, 2.4)

Definition 18. Let G = ( N, M) be a neutrosophic graph. G is said to be a (m, (d1 , d2 , d3 ))-regular

neutrosophic graph or dm -regular if for all nodes a ∈ V in G, dm ( a) = (d1 , d2 , d3 ).

Example 3. An example of an (m, (d1 , d2 , d3 ))-totally regular neutrosophic graph is pictured in Figure 2.
Figure 2 is a (2, (0.1, 0.6,0.9))-regular neutrosophic graph.

(.2, .3, .5) (.3, .5, .7)

a b
(.1, .7, .9)
(.2, .6, .9)

(.2, .6, .8)

c d
(.1, .7, .9)
(.5, .6, .8) (.4, .6, .5)

Figure 2. A neutrosophic graph G.

Mathematics 2019, 7, 551 7 of 20

d2 ( a) = (((0.2 + 0.2) + 0.3) , ((0.6 + 0.8) + 0.6) , ((0.8 + 0.9) + 0.7)) = (0.7, 2.0, 2.4)
td2 (b) = (((0.2 + 0.2) + 0.3) , ((0.8 + 0.8) + 0.6) , ((0.9 + 0.9) + 0.6)) = (0.7, 2.1, 2.4)
td2 ( a) = (((0.2 + 0.2) + 0.4) , ((0.8 + 0.8) + 0.5) , ((0.9 + 0.9) + 0.6)) = (0.8, 2.1, 2.4)

Definition 19. Let G = ( N, M) is a neutrosophic graph. G is an (m, (k1 , k2 , k3 ))-totally regular neutrosophic
graph or tdm -regular neutrosophic graph if for all nodes a ∈ V in G, tdm ( a) = (k1 , k2 , k3 ).

Example 4. An example of a (m, (k1 , k2 , k3 ))-totally regular neutrosophic graph is shown in Figure 3. It is a
(2, (0.6, 2.0,2.1))-regular neutrosophic graph.

(.3, .4, .5) (.4, .5, .6)

a b
(.3, .7, .8)
(.1, .6, .8)

)
, .7
., 6
(.2

c
z(.3, .5, .5)

Figure 3. A neutrosophic graph G.

Theorem 1. Let G = ( N, M ) be an SVNG. If TN , IN and FN are constant functions, then G is a dm -regular

neutrosophic graph if G is a tdm -regular neutrosophic graph (m is a positive integer).

Proof. Suppose that for every node v in G, ( TN (v) , IN (v) , FN (v)) = (c1 , c2 , c3 ) and
dm (v) = (d1 , d2 , d3 ). Then:

tdm (v) = dm (v) + ( TN (v) , IN (v) , FN (v)) = (d1 + c1 , d2 + c2 , d3 + c3 ) ;

hence, G is a tdm -regular neutrosophic graph. If G is a tdm -regular neutrosophic graph, then the proof
is similar to the previous case.

Theorem 2. Let G = ( N, M ) be an (m, (d1 , d2 , d3 ))-totally regular and an (m, (k1 , k2 , k3 ))-totally
regular neutrosophic graph with n nodes. Then, TN , IN and FN are constant functions, and
O ( G ) = n ( k 1 − d1 , k 2 − d2 , k 3 − d3 ).

Proof. If G is an (m, (d1, d2, d3))-regular neutrosophic graph and an (m, (k1, k2, k3))-totally regular neutrosophic
graph respectively, then for all v ∈ V we get,

dm (u) = (d1, d2, d3) → ( ∑ m

TM (uv), ∑ m
Ib (uv), ∑ m
Fb (uv)) = (d1, d2, d3)
6 v∈V
u= 6 v∈V
u= 6 v∈V
u=

and:

tdm (v) = (k1, k2, k3) → ( ∑ m

TB (uv), + TN (v), ∑ m
IM (uv), + IN (v), ∑ m
FM (uv), + FN (v)) = (k1, k2, k3).
6 v∈V
u= 6 v∈V
u= 6 v∈V
u=
Mathematics 2019, 7, 551 8 of 20

Hence, by (1), ( d1 , d2 , d3 ) = (k1 − TN (v) , k2 − IN (v) , k3 − FN (v)) and so

( TN (v) , IN (v) , FN (v)) = (k1 − d1 , k2 − d2 , k3 − d3 ). Then, TN , IN and FN are constant functions, and
since G has n nodes, we get:

O( G ) = ( ∑ TN (v), ∑ I N ( v ), ∑ FN (v)) = n(k1 − d1 , k2 − d2 , k3 − d3 ).

v ∈V v ∈V v ∈V

Free and Busy Vertex on the Regular Neutrosophic Graph

Definition 20. Let G = ( N, M) be a regular neutrosophic graph and a ∈ V be a node of G. The neutrosophic busy
value of a a is denoted by D(a) = (DT (a) , DI (a) , DF (a)), where:

DT (a) = ∑ min (TN (a) , TN (ai )) , DI (a) = ∑ max ( IN (a) , IN (ai )) , DF (a) = ∑ max (FN (a) ∨ FN (ai ))

Here, ai denotes the adjacent vertex of a of G. The busy value of regular neutrosophic graph G is the sum of
all busy values of all the nodes of G.

Definition 21. Let G = ( N, M ) be a regular neutrosophic graph and a ∈ V be a node of G. The a of G is a

busy node or vertex if:
TN ( a) ≤ d T ( a) , IN ( a) ≤ d I ( a) , FN ( a) ≥ d F ( a)

If the vertex a is not a busy vertex, then it is said to be a free vertex.

Definition 22. Let G = ( N, M) be a regular neutrosophic graph and ( a, b) ∈ E be an arc of G. a, b is said to

be an effective edge if:

TM ( a, b) = min ( TN ( a) , TN (b)) , I M ( a, b) = max ( IN ( a) , IN (b)) , FM ( a, b) = max ( FN ( a) , FN (b))

4. Regularity on the Complement and Isomorphic Neutrosophic Graph

Definition 23. Let G1 = ( N1 , M1 ) and G2 = ( N2 , M2 ) be two regular neutrosophic graphs.

1. A homomorphism h from a regular neutrosophic graph G1 to another regular neutrosophic graph G2 is a

neutrosophic mapping function h : V1 → V2 , which always satisfies following conditions:

(i ) ∀ a1 ∈ V1 , a1 b1 ∈ E1
( a) TN1 ( a1 ) ≤ TN2 (h ( a1 )) , IN1 ( a1 ) ≤ IN2 (h ( a1 )) , FN1 ( a1 ) ≥ FN2 (h ( a1 ))

(b) TB1 ( a1 b1 ) ≤ TM2 (h ( a1 ) h ( a2 )) , I M1 ( a1 b1 ) ≤ I M2 (h ( a1 ) h ( a2 )) ,

FM1 ( a1 b1 ) ≥ FB2 (h ( a1 ) h ( a2 ))
2. A isomorphism h from a regular neutrosophic graph G1 to another regular neutrosophic graph G2 is a
neutrosophic bijective mapping function h : V1 → V2 , which always satisfies the following:

(i ) ∀ a1 ∈ V1 , a1 b1 ∈ E1
(c) TN1 ( a1 ) = TN2 (h ( a1 )) , IN1 ( a1 ) = IN2 (h ( a1 )) , FN1 ( a1 ) = FN2 (h ( a1 ))

(d) TM1 ( a1 b1 ) = TM2 (h ( a1 ) h ( a2 )) , I M1 ( a1 b1 ) = I M2 (h ( a1 ) h ( a2 )) ,

FM1 ( a1 b1 ) = FM2 (h ( a1 ) h ( a2 ))
3. A weak isomorphism h from a regular neutrosophic graph G1 to another regular neutrosophic graph G2 is
a neutrosophic bijective mapping function h : V1 → V2 , which always satisfies the following:

(i ) ∀ a1 ∈ V1
Mathematics 2019, 7, 551 9 of 20

(e) h is a homomorphism.

(f) TN1 ( a1 ) = TN2 (h ( a1 )) , IN1 ( a1 ) = IN2 (h ( a1 )) , FN1 ( a1 ) = FN2 (h ( a1 )).

4. A co-weak isomorphism h from a regular neutrosophic graph G1 to another regular neutrosophic graph G2
is a neutrosophic bijective mapping function h : V1 → V2 , which always satisfies the following:

(i ) ∀ a1 b1 ∈ E1
( g) h is a homomorphism.

(h) TN1 ( a1 b1 ) = TN2 (h ( a1 b1 )) , IN1 ( a1 b1 ) = IN2 (h ( a1 b1 )) , FN1 ( a1 b1 ) = FN2 (h ( a1 b1 )).

Definition 24. Let G = ( N, M ) be a connected neutrosophic graph. The neutrosophic graph G is called an
HING if and only if all the adjacent nodes of G are always different neutrosophic degrees.

Theorem 3. If G1 and G2 are two isomorphic highly-irregular neutrosophic graphs (HING), then their size and
order are always equal.

Proof. An isomorphism h is between two HINGs G1 and G2 with the underlying node set V1 and V2 ,
respectively, then:

(i ) ∀ a1 ∈ V1 , a1 b1 ∈ E1

( a) TN1 ( a1 ) = TN2 (h ( a1 )) , IN1 ( a1 ) = IN2 (h ( a1 )) , FN1 ( a1 ) = FN2 (h ( a1 ))

(b) TM1 ( a1 b1 ) = TB2 (h ( a1 ) h ( a2 )) , IM1 ( a1 b1 ) = I M2 (h ( a1 ) h ( a2 )) ,

FM1 ( a1 b1 ) = FM2 (h ( a1 ) h ( a2 ))

Therefore, we have:
!
S( G1 ) = ∑ TM1 ( a1 b1 ) , ∑ I M1 ( a1 b1 ) , ∑ FM1 ( a1 b1 )
a1 b1 ∈ E1 a1 b1 ∈ E1 a1 b1 ∈ E1
!
= ∑ TM2 (h( a1 ), h(b1 )) , ∑ I M2 (h( a1 ), h(b1 )) , ∑ FM2 (h( a1 ), h(b1 ))
a1 ,b1 ∈V1 a1 ,b1 ∈V1 a1 b1 ∈V1
!
= ∑ TM2 ( a2 b2 ) , ∑ I M2 ( a2 b2 ) , ∑ FM2 ( a2 b2 ) = S( G2 )
a2 b2 ∈ E2 a2 b2 ∈ E2 a2 b2 ∈ E2

!
O( G1 ) = ∑ TN1 ( a1 ) , ∑ IN1 ( a1 ) , ∑ FN1 ( a1 )
a1 ∈V1 a1 ∈V1 a1 ∈V1
!
= ∑ TN2 (h( a1 )) , ∑ IN2 (h( a1 )) , ∑ FN2 (h( a1 ))
a1 ∈V1 a1 ∈V1 a1 ∈V1
!
= ∑ TN2 ( a2 ) , ∑ IN2 ( a2 ) , ∑ FN2 ( a2 ) = O( G2 )
a2 ∈V2 a2 ∈V2 a2 ∈V2

Theorem 4. Let G1 and G2 be two isomorphic HINGs. Then, the degrees of nodes a and h( a) of G1
and G2 are the same.
Mathematics 2019, 7, 551 10 of 20

Proof. If G1 and G2 are two isomorphic HINGs, respectively, then TM1 ( a) = TM1 (h( a1 )h(b1 )),
I M1 ( a) = I M1 (h( a1 )h(b1 )) and FM1 ( a) = FM1 (h( a1 )h(b1 )) for all a1 , b1 ∈ V1 . Therefore,

d T ( a1 ) = ∑ TM1 ( a1 b1 ) = ∑ TM2 (h( a1 )h(b1 )) = d T (h( a1 ))

a1 ,b1 ∈V1 a1 ,b1 ∈V1

d I ( a1 ) = ∑ I M1 ( a1 b1 ) = ∑ I M2 (h( a1 )h(b1 )) = d I (h( a1 ))

a1 ,b1 ∈V1 a1 ,b1 ∈V1

d F ( a1 ) = ∑ FM1 ( a1 b1 ) = ∑ FM2 (h( a1 )h(b1 )) = d F (h( a1 ))

a1 ,b1 ∈V1 a1 ,b1 ∈V1

for all a1 ∈ V1 . This proves that the degrees of nodes u and h(u) of G1 and G2 are preserved.

Definition 25. Let G = ( N, M) be a connected neutrosophic graph. G is defined as: mm

1. A self-weak neutrosophic complementary isomorphic graph if graph G is weak isomorphic with G c .

2. A self-complementary neutrosophic graph if G c ∼
= G.

Theorem 5. Let G be a self-complementary HING. Then:

1
∑ TM (a, b) ≤ 2 ∑ min (TN (a) , TN (b)))
a6=b a6=b
1
∑ IM (a, b) ≥ 2 ∑ max ( IN (a) , IN (b)))
a6=b a6=b
1
∑ FM (a, b) ≥ 2 ∑ max ( FN (a) , FN (b)))
a6=b a6=b

Definition 26. Let G = ( N, M) be a neutrosophic

 graph and G µ = ( N µ , Mµ ) be the µ complement of G
µ µ µ µ µ
where N = N and M = TM , I M , FM where:
(
µ min ( TN ( a), TM (b)) − TM ( a, b) if TM ( a, b) > 0
TM =
0 if TM ( a, b) = 0
(
µ I M ( a, b) − max ( TN ( a), TM (b)) if I M ( a, b) > 0
IM =
0 if I M ( a, b) = 0
(
µ FM ( a, b) − max ( FN ( a), FM (b)) if FM ( a, b) > 0
FM =
0 if FM ( a, b) = 0

Theorem 6. Let G µ be the µ-complement of an HING. G need not be highly irregular.

Proof. All the adjacent nodes of G with different neutrosophic degrees or all the other nodes of G with
different neutrosophic degrees may possibly be adjacent nodes with equivalent degrees. This condition
opposes the idea of HING.

µ
Theorem 7. Let G1 and G2 be two HINGs. If G1 and G2 are two isomorphic neutrosophic graphs, then G1 and
µ
G2 are also isomorphic, and vice versa.

Proof. Let G1 and G2 be two isomorphic neutrosophic graphs; there always exists a map h : V1 toV2
satisfying TN1 (a) = TN2 (h (a)), for all a ∈ V1 and TM1 (ab) = TM2 (h (a) h (b)), IM1 (ab) = IM2 (h (a) h (b))
and FM1 (ab) = FM2 (h (a) h (b)), for all ab ∈ E1. By the concept of the µ complement neutrosophic graph,
µ
we have TM1 (ab) = min (TN1 (a) , TN1 (b) − TM1 (ab) = min (TN2 (h (a)) , TN2 (h (b)))) − TM2 (h (a) h (b)),
µ
FM1 (ab) = FM1 (ab) − max (FN1 (a) , FN1 (b)) = TM2 (h (a) h (b)) − max (FN1 (h (a)) , FN2 (h (b))) for all
Mathematics 2019, 7, 551 11 of 20

ab ∈ E1 (IM1 (ab) values can be calculated using the same method). Hence, G1 ∼
µ µ µ
= G2 . The proof of the
converse part is straight forward.

5. Complete Bipartite Neutrosophic Graphs

Definition 27. A neutrosophic graph G = ( N, M ) is called a complete bipartite neutrosophic graph if the
vertex set V can be divided into two nonempty sets, such that for every v1 , v2 ∈ V1 or V2 and for every
u ∈ V1 and v ∈ V2 :

Theorem 8. Let G = ( N, M) be a complete bipartite neutrosophic graph, such that each partition of V has
nodes.
(i) If for every v ∈ V1 \V2 or v ∈ V2 \V1 , (t N (v), i N (v), f N (v)) = (c1 , c2 , c3 ) and for the other nodes
as u, t N (u) ≥ c1 , i N (u) ≤ c2 , f N (u) ≤ c3 , , then for all 1 ≤ m ≤ 2k − 1, Gis a dm regular
neutrosophic graph.
(ii) Let W = { v1 , v2 , ..., vk } be a set of nodes, such that it contains at least one node of each partition of V.

Proof. Suppose that for every v1 ∈ V1\V2 , (tN (v), iN (v), f N (v)) = (c1 , c2, c3). Therefore, in V1, tN and f N
are constant functions, and since for every v2 ∈ V1\V2 , tN (v1) ≤ tN (v1) , iN (v1) ≥ iN (v2), f N (v1) ≥
f N (v21) ). Then, by the previous definition, we have that for every uv ∈ E, we get (tM (uv), iM (uv), f M (uv)) =
(c1 , c2 , c3 ).
Hence, for every v ∈ V and 1 ≤ m ≤ 2k − 1, if m is odd, then dm (v) = k(c1 , c2 , c3 ) and if m is
even, then dm (v) = (k − 1)(c1 , c2 , c3 ).
Therefore, in both cases, we see that G is a dm -regular neutrosophic graph. Now, if for every
v1 ∈ V1 \V2 , (t N (v), i N (v), f N (v)) = (c1 , c2 , c3 ), then the proof is similar to the proof of (i).
(ii) The proof is similar to the previous theorem only by the difference, that every v ∈ V and k +
1 ≤ m ≤ 2k − 1; if m is odd, then and if m is even, then dm (v) = (k − 1)(c1 , c2 , c3 ).

Remark 1. If V1 = { v1 , v2 , ..., vn } and V1 = { u1 , u2 , ..., um } such that n 6= m, , then the above theorem
does not hold.

Example 5. Consider the complete bipartite neutrosophic graph G = ( N, M ) in Figure 4 shown below.

(.2, .3, .6) (.2, .3, .6) z(.2, .3, .6)

a b c

e d
(.2, .3, .6) (.2, .3, .6)

Figure 4. An example of a complete bipartite neutrosophic graph.

Since G is a complete bipartite neutrosophic graph, then for every

uv ∈ E, (tM(uv), iM(uv), f M(uv)) = (0.2, 0.3, 0.6). Hence:

d1(a) = (0.4, 0.6, 1.2), d1(b) = (0.4, 0.6, 1.2), d1(c) = (0.4, 0.6, 1.2), d1(d) = (0.6, 0.9, 1.8), d1(e) = (0.6, 0.9, 1.8)
Mathematics 2019, 7, 551 12 of 20

d2(a) = (0.4, 0.6, 1.2), d2(b) = (0.4, 0.6, 1.2), d2(c) = (0.4, 0.6, 1.2),

d2(d) = (0.2, 0.3, 0.6), d2(e) = (0.2, 0.3, 0.6)

d3(a) = (0.4, 0.6, 1.2), d3(b) = (0.4, 0.6, 1.2), d3(c) = (0.4, 0.6, 1.2),

d3(d) = (0.6, 0.9, 1.8), d3(e) = (0.6, 0.9, 1.8)

d4(a) = (0.4, 0.6, 1.2), d4(b) = (0.4, 0.6, 1.2), d4(c) = (0.4, 0.6, 1.2),

d4(d) = (0, 1, 1), d4(e) = (0, 1, 1)

Since t N , i N and f N are constant functions, we have for all m = 1, 2, 3, 4, G does not follow the condition
of the dm -regular neutrosophic graph.

Theorem 9. Let G = ( N, M ) be a regular neutrosophic graph on cycle graph G ∗ = (V, E) with k ≥ 3 nodes.
If for k − (n − 1) arcs of G as uv, ( TM (uv), I M (uv), FM (uv)) = (c1 , c2 , c3 ) and for arcs as ab, TM ( ab) ≥ c1 ,
I M ( ab) ≥ c2 , and FM ( ab) ≥ c3 , then for any n ≤ m ≤ k − 1, G is a dm -regular neutrosophic graph.

Proof. Let G = ( N, M ) be a regular neutrosophic graph on cycle graph G ∗ with k nodes and for
k − (n − 1) arcs of G as uv, ( TM (uv), I M (uv), FM (uv)) = (c1 , c2 , c3 ).
Since G ∗ is a cycle, we have for every node w in G that there exist exactly two distinct nodes w1
and w2 such that w and w1 are connected by means of one path of length m (where n ≤ m ≤ k − 1)
and each path contains at least one arc uv.
Hence, for two paths,
λ1 : w1 = u0 , u1 , u2 , ...um−1 , um = w and λ2 : w1 = v0 , v1 , v2 , ...vm−1 , vm = w.

Therefore, we obtain
m ( w w ) + T m ( w w ) = T ( w u ) ∧ ... ∧ T ( u
TM 1 M 2 M 1 1 M m−1 w ) + TM ( w2 v1 ) ∧ ... ∧ TM ( vm−1 w ) = c1 + c1 = 2c1 .
m m
I M (w1 w) + I M (w2 w) = I M (w1 u1 ) ∨ ... ∨ I M (um−1 w) + I M (w2 v1 ) ∨ ... ∨ I M (vm−1 w) = c1 + c1 = 2c1 .
m ( w w ) + F m ( w w ) = F ( w u ) ∨ ... ∨ F ( u
FM 1 M 2 M 1 1 M m−1 w ) + FM ( w2 v1 ) ∨ ... ∨ FM ( vm−1 w ) = c1 + c1 = 2c1 .

Remark 2. Let G be a regular neutrosophic graph on cycle graph G ∗ with k nodes. If TM , I M , and FM are
constant functions and TN , IN , and FN are not, then for any 1 ≤ m ≤ k − 1, G is not a tdm -regular neutrosophic
graph in general.

Theorem 10. Let G = ( N, M ) be a neutrosophic graph on cycle graph G ∗ = (V, E) with k ≥ 3 nodes. If G is
a d1 -regular neutrosophic graph, then TM , I M , and FM are constant functions.

Proof. Suppose that G is a d1 -regular neutrosophic graph and there exists uv ∈ E

such that ( TM (uv), I M (uv), FM (uv)) = (c1 , c2 , c3 ) , where for every ab ∈ E − {uv},
( TM ( ab), I M ( ab), FM ( ab)) = (c1 , c2 , c3 ). Then, TM (uv) 6= c1 , IM (uv) 6= c2 , and FM (uv) 6= c3 .
Case 1 : If TM (uv) 6= c1 ,
∑ TM1 (uz) = c1 + TM (uv), ∑ TM1 (vz) = c1 + TM (uv).
u 6 = z ∈V v 6 = z ∈V
Since G has k ≥ 3 nodes, there exists w ∈ V such that w 6= v, u. Hence, we get ∑ 1
TM (wz) =
u 6 = z ∈V
2c1 , which implies that TM1 ( wz ) 6 = T 1 ( vz ) and T 1 ( wz ) 6 = T 1 ( uz ). Therefore, d ( w ) 6 = d ( u ) and
M M M 1 1
d1 (w) 6= d1 (v); hence, G is not a d1 -regular neutrosophic graph, which is a contradiction.
Mathematics 2019, 7, 551 13 of 20

Case 2 : If I M (uv) 6= c2 ,
∑ 1
IM (uz) = c2 + I M (uv), ∑ 1
IM (vz) = c2 + IM (uv).
u 6 = z ∈V v 6 = z ∈V
Since G has k ≥ 3 nodes, there exists w ∈ V such that w 6= v, u. Hence, we get ∑ 1
IM (wz) =
u 6 = z ∈V
2c2 , which implies that I M 1 ( wz ) 6 = I 1 ( vz ) and I 1 ( wz ) 6 = I 1 ( uz ). Therefore, d ( w ) 6 = d ( u ) and
M M M 1 1
d1 (w) 6= d1 (v); hence, G is not a d1 -regular neutrosophic graph, which is a contradiction.
Case 3 : If FM (uv) 6= c3 ,
∑ FM1 (uz) = c3 + FM (uv), ∑ FM1 (vz) = c3 + FM (uv).
u 6 = z ∈V v 6 = z ∈V
Since G has k ≥ 3 nodes, there exists w ∈ V such that w 6= v, u. Hence, we get
∑ FM1 1 ( wz ) 6 = F1 ( vz ) and F1 ( wz ) 6 = F1 ( uz ). Therefore,
(wz) = 2c3 , which implies that FM M M M
u 6 = z ∈V
d1 (w) 6= d1 (u) and d1 (w) 6= d1 (v); hence, G is not a d1 -regular neutrosophic graph, which is a
contradiction. Hence the proof.

Theorem 11. Let G = ( N, M ) be a neutrosophic graph on the path graph G ∗ = (V, E) with 2m nodes. If for
at least m + 1 arcs of G as uv, ( TM (uv), I M (uv), FM (uv)) = (c1 , c2 , c3 ) and for other arcs as ab, TM ≥ c1 ,
I M ≤ c2 and FM ≤ c3 , then G is also a dm -regular neutrosophic graph.

Proof. Suppose that G ∗ is a path with 2m nodes and for at least m + 1arcs of G as uv,
( TM (uv), IM (uv), FM (uv)) = (c1 , c2 , c3 ). Since G ∗ is a path, we have for every arbitrary node w
in G that there exists only one node as z such that w and z are connected by means of one path of
length m and this path contains at least one uv. Hence, for the path λ : z = z0 , z1 , z2 , ...zm−1 , zm = w.
m ( zw ) = T ( zz ) ∧ ... ∧ T ( z m
We get TM M 1 M m−1 w ) = c1 , I M ( zw ) = I M ( zz1 ) ∨ ... ∨ I M ( zm−1 w ) = c2 , and
m
FM (zw) = FM (zz1 ) ∨ ... ∨ FM (zm−1 w) = c3 .
Therefore, dm (w) = ( TM m ( zw ), I m ( zw ), F m ( zw )) = (c1 , c2 , c3 ), and so, G is a
M M
(m, (c1 , c2 , c3 ))-regular neutrosophic graph.

Remark 3. Let G = ( N, M ) be a neutrosophic graph on the path graph G ∗ = (V, E) with 2m nodes. If TM ,
I M , and FM are constant functions and TN , IN , and FN are not, then G is not tdm -regular.

6. m-Highly Irregular Neutrosophic Graph

In this section, we present the notation of m-HING and the m-highly totally irregular neutrosophic
graph (HTING). We have proven the necessary and sufficient conditions for which the m-highly
irregular and m-HTINGs are equivalent, and also, we provide m-highly irregularity on neutrosophic
graph G where G m is the cycle, path, or star graph.

Definition 28. Let G = ( N, M) be a regular connected neutrosophic graph, then:

(i ) G is said to be m-HING, if each node of G is adjacent to the other nodes with the different values of the
dm -degree.
(ii ) G is said to be m-HTING, if each node of G is adjacent to the other nodes with the different values of the
tdm -degree.

Theorem 12. Let G = ( N, M ) be a neutrosophic graph. If TM , I M , and FM are used to denote the three
constant functions and m represents a positive integer number, then G is m-HTING if G is m-HING.

Proof. Suppose that G is m-HTING. Then, the value of the tdm -degree of all the pairs of adjacent nodes
are different. Let two nodes v and u be the pair of adjacent nodes with the distinct tdm -degree. Hence,
we have
Mathematics 2019, 7, 551 14 of 20

" #
tdm (u) = ∑ m
TM (uw) + TN (u), ∑ m
IM (uw) + IN (u), ∑ m
FM (uw) + FN (u)
u 6 = w ∈V u 6 = w ∈V u 6 = w ∈V
" #
tdm (v) = ∑ m
TM (vw) + TN (v), ∑ m
IM (vw) + IN (v), ∑ m
FM (vw) + FN (v) .
v 6 = w ∈V v 6 = w ∈V v 6 = w ∈V
Since tdm (u) 6= tdm (v), we have, ∑ m
TM (uw) + TN (u) 6= ∑ m
TM (vw) + TN (v), ∑ m
IM (uw) +
u6=w∈V v6=w∈V u6=w∈V
IN (u) 6= ∑ m
IM (vw) + IN (v) and ∑ m
FM (uw) + FN (u) 6= ∑ m
FM (vw) + FN (v).
v6=w∈V u6=w∈V v6=w∈V
Since TN (u) = TN (v), IN (u) = IN (v), and FN (u) = FN (v), hence we get
∑ TMm (uw) 6= ∑ TMm (vw), ∑ IMm (uw) 6= ∑ IMm (vw), and ∑ FMm (uw) 6= ∑ m
FM (vw).
u6=w∈V v6=w∈V u6=w∈V v6=w∈V u6=w∈V v6=w∈V
Therefore,"we have, # "
dm (u) = ∑ m
TM (uw), ∑ m
IM (uw), ∑ m
FM (uw) 6= ∑ m
TM (vw),
u6=w∈V u6=w∈V u6=w∈V v6=w∈V
#
∑ m
IM (vw), ∑ m
FM (vw) = dm (v). Hence, any pair of adjacent nodes in G has distinct a dm -degree.
v6=w∈V v6=w∈V
Then, G is m-HING.
Conversely, G is an m-HING. Then, the value of the dm-degree of each and every pair of adjacent nodes
6 dm (v). Since TN (u) = TN (v), IN (u) = IN (v), and
such as u and v is distinct. This implies that dm (u) =
FN (u) = FN (v), hence tdm (u) = dm (u) + TN (u) = 6 idm (u) = dm (u) + IN (u) =
6 f dm (u) = dm (u) + FN (u),
and so, any two adjacent nodes in G has the distinct tdm-degree. Therefore, G is an m-HTING.
Hence the proof.

Theorem 13. Let G = ( N, M ) be a neutrosophic graph on cycle graph G ∗ = (V, E) with k ≥ 3 nodes and
for all i = 1, 2, ..., k − 1 (where vk+1 = v1 ),"TM (vi vi+1 )# < TM (vi+1 vi+2 )or I M (vi vi+1 ) > I M (vi+1 vi+2 ) or

FM (vi vi+1 ) > FM (vi+1 vi+2 ). Then, G is a 1, 2, 3, 2k -HING.

 

Proof. Suppose that v1 , v2 , ..., vk , v1 is the arcs of G, where TM (v1 v2 ) < TM (v2 v3 ) < ... < TM (vi vi+1 ) <
TM (vi+1 vi+2 ) < ... < TM (vk v1 ) and there exists vi vi+1 ∈ E such that for 1 ≤ m ≤ 2k , dm (vi ) =
 

dm (vi+1 ). Therefore, we get ∑ TMm (vi v j ) = TM (vi vi+1 ) + TM (v j v j+1 ) and ∑ TMm (vi v j ) =
v i 6 = v i +1 ∈ V v i +1 6 = v j ∈ V
TM (vi vi+1 ) + TM (v j+1 v j+2 ). Since dm (vi ) = dm (vi+1 ), TM (vi vi+1 ) + TM (v j v j+1 ) = TM (vi vi+1 ) +
TM (v j+1 v j+2 ), thus TM (vi vi + 1) = TM (v j+1 vj + 2), which is a contradiction.
Now, for every v1 , v2 , ..., vk , v1 ∈ E, also we have
I M (v1 v2 ) > I M (v2 v3 ) > ... > I M (vi vi+1 ) > I M (vi+1 vi+2 ) > ... > I M (vk v1 ),
FM (v1 v2 ) > FM (v2 v3 ) > ... > FM (vi vi+1 ) > FM (vi+1 vi+2 ) > ... > FM (vk v1 ).
The" proof is similar
# to the case as true membership, and the contradiction is obtained. Therefore,
k
G is an 1, 2, 3, 2 -HING.

Remark 4. Let G = ( N, M ) be a neutrosophic graph on cycle graph G ∗ = (V, E) with k ≥ 3 nodes and
for all i = 1, 2, ..., k − 1 (where vk+1 = v1 ), TM"(vi vi+1 ) < #TM (vi+1 vi+2 )or I M (vi vi+1 ) > I M (vi+1 vi+2 ) or

FM (vi vi+1 ) > FM (vi+1 vi+2 ). Then, G is not a 1, 2, 3, 2k -HING in general.

 

Example 6. Consider the neutrosophic graph G = ( N, M ) on graph G ∗ = (V, E) shown in the Figure 5.
Mathematics 2019, 7, 551 15 of 20

(.4, .5, .7) (.3, .6, .7)

a b
(.1, .7, .8)

(.4, .6, .9)

(.2, .7, .8)

(.3, .6, .8)
d c
(.5, .5, .7) (.4, .5, .8)

Figure 5. m-highly irregular neutrosophic graph.

We see that TM ( ab) < TM (bc) < TB (cd) < TB (da), and we get
td2 ( a) = (0.3 + 0.4, 0.6 + 0.4) = (0.7, 1.0)
td2 (b) = (0.2 + 0.3, 0.6 + 0.6) = (0.5, 1.2)
td2 (c) = (0.3 + 0.4, 0.6 + 0.5) = (0.7, 1.1)
td2 (d) = (0.2 + 0.5, 0.6 + 0.4) = (0.7, 1.0)
Since for two vertices a and b, they are adjacent, td2 ( a) = td2 (b). Then, G is not a two-HTING.

Theorem 14. Let G = ( N, M ) be a neutrosophic graph on cycle graph G ∗ = (V, E) with k ≥ 3 nodes and
for all i = 1, 2, ..., k − 1 (where vk+1 = v1 ),"TM (vi vi+1 )# < TM (vi+1 vi+2 )or I M (vi vi+1 ) > I M (vi+1 vi+2 ) or

FM (vi vi+1 ) > FM (vi+1 vi+2 ). Then, G is a 1, 2, 3, 2k -HING.

 

Proof. Suppose that v1 , v2 , ..., vk , v1 is the arcs of G, where TM (v1 v2 ) < TM (v2 v3 ) < ... < TM (vi vi+1 ) <
TM (vi+1 vi+2 ) < ... < TM (vk−1 vk ). Then, for every 1 ≤ m ≤ 2k , we get
 


 TM (vi vi+1 ), i f i = 1, 2, 3, ..., m

∑ TMm (vi v j ) =  TM (vi vi+1 ) + TM (vi−m vi−m+11 , i f i = m + 1, ..., k − m
vi 6 = v j ∈V 
TM (vi−m vi−m+1 , i f i = k − m + 1, ...k.


" Hence, for every

# two adjacent nodes as vi and vi+1 in G, we get dm (vi ) 6= dm (vi+1 ). Therefore, G

is a 1, 2, 3, 2 − 1 -HING. Now, suppose that m = 2k . If k is even, then we have

k   

(
TM (vi vi+1 ), i f i = 1, 2, 3, ..., m
∑ m
TM ( vi v j ) =
vi 6 = v j ∈V TM (vi−m v ), i f i = m + 1, ..., k − m
If k is odd, thenwe have

 TM (vi vi+1 ), i f i = 1, 2, 3, ..., m

∑ TMm (vi v j ) =  TM (v1 v2 ) + TM (vi−m vi−m+11 , i f i = m + 1
vi 6 = v j ∈V 

TM (vi−m vi−m+1 , i f i = m + 2, ...k.
Here, in both cases, we see that for every two adjacent nodes vi and vi+1 in G, we get
dm (vi ) 6= dm (vi+1 ), and so, G is a 2k -HING. Similarly, for the indeterministic and false membership
 

function, for each arc, it is proven.

Hence the proof.

Example 7. Consider the the neutrosophic graph G = ( N, M ) on graph G ∗ = (V, E) shown in Figure 6.
Mathematics 2019, 7, 551 16 of 20

(.4, .4, .7) (.3, .3, .8)

a b
(.1, .7, .8)

(.4, .6, .9)

(.2, .7, .8)

d c
(.5, .5, .7) (.4, .5, .8)

Figure 6. Neutrosophic graph.

We see that TM ( ab) < TM (bc) < TM (cd), and we get

td2 ( a) = (0.1 + 0.4, 0.4 + 0.4) = (0.5, 0.8)
td2 (b) = (0.2 + 0.3, 0.5 + 0.3) = (0.5, 0.8)
td2 (c) = (0.1 + 0.4, 0.4 + 0.2) = (0.5, 0.6)
td2 (d) = (0.2 + 0.5, 0.5 + 0.4) = (0.7, 0.9)
Since for two adjacent nodes a and b, td2 ( a) = td2 (b). Then, we see that G is not a two-HTING.

Definition 29. Let G = ( N, M ) be a neutrosophic graph. The graph G is said to be a star neutrosophic graph
if it is a complete bipartite neutrosophic graph such that one partition of V contains only one node.

Theorem 15. Let G = ( N, M ) be a star neutrosophic graph with k ≥ 4 nodes such that the node v1 is incident
to the other nodes. If for all i = 1, 2, ..., k, TM (vi vi+1 ) < TM (vi+1 vi+2 ), or I M (vi vi+1 ) > I M (vi+1 vi+2 ), or
 
FM (vi vi+1 ) > FM (vi+1 vi+2 ). Then, G is a 1, 2 -HING.

Proof. Suppose that v1 , v2 , ..., vk , v1 is the arcs of G, where TM (v1 v2 ) < TM (v2 v3 ) < ... < TM (vi vi+1 ) <
TM (vi+1 vi+2 ) < ... < TM (vk v1 ).
k
Then, ∑ 1
TM ( v1 v j ) = ∑ TM (v1 v j ) and ∑ 1
TM (v1 v j ) = TM (v1 v j ); i = 2, 3, ..., k.
vi 6 = v j ∈V j =2 vi 6 = v j ∈V
Hence, d1 (v1 ) 6= d1 (v2 ) 6= d1 (v3 ) 6= .... 6= d1 (vk ). There, G is a one-HING. In addition, we see that
i −1
d1 (v1 ) = (0, 1) and ∑ 2
TM ( v1 v j ) = ( k − i ) t M ( v1 v i ) + ∑ TM (v1 v j ). Hence, for every i = 2, ..., k,
vi 6 = v j ∈V j =2
d1 (vi ) 6= d1 (v1 ), and so, G is a two-HING. Similarly, the proof is established for the indeterminacy and
false membership function of each arc.
Hence the proof.

7. Some Applications of Regular and Bipartite Neutrosophic Graphs

There are many networks that are used in real-life scenarios. Graph theory plays a vital role in
the application parts of these networks. Some networks are designed for distributing commodities or
services. The power grid, Internet-related issues, and road transportation networks are some examples
of technological networks. There has been a phenomenal change in network research in recent years
from a small graphical structure to a wide statistical analysis of large-scale networks.
Most of the real-life networks/structures are ambiguous and vague in nature and not well defined.
Indeterministic information exists in almost every real-life scenario. The neutrosophic graph is more
flexible and compatible to model the indeterministic information in a network when compared to the
fuzzy graph or vague graph. A graph is a classical way of modeling real-life networks, which consist of
relationships between entities. In a graph, the entities are expressed by nodes and relationships by arcs.
Mathematics 2019, 7, 551 17 of 20

If there exist uncertainties due to indeterministic or vague information about the entities, or relations,
or both, then the neutrosophic graph model is very efficient to design such networks. Neutrosophic
graphs are widespread and important in the area of mathematical modeling, pattern recognition,
relational mapping, biological, social networks, and information systems. It can be applied to artificial
intelligence with neutrosophic rules in the inference system. These graphs can be represented in
different data structures in different applications internally. In those real-life applications, the basic
mathematical model is that of a neutrosophic graph.
Transforming all those data-carrying networks to a neutrosophic graph using various types
of information from all the possible existing levels, it is a proper multilevel method where all
possible levels and ties among them are represented and analyzed mathematically at the same time.
Neutrosophic graphs have more applications related to information technology and computer science
such that these graphs are used to illustrate networks of communication, data optimization, machine
learning, chip design, and much more.

7.1. Assignment Problem

The assignment problem is a well-known and fundamental topic of study in the field of operation
research. This problem can be modeled as a bipartite neutrosophic graph. Let a neutrosophic graph
G = (V, E) where V denotes the nodes and E represents the arcs be represented as a neutrosophic
bipartite graph, if it satisfies the condition such that V = X ∪ Y and X ∩ Y = v; where v is a null
matrix, and E ∈ X × Y. A neutrosophic bipartite graph G is described as a weighted bipartite graph if
every edge (i.e., E) of it consists of a neutrosophic weight (i.e., v (i, j)). The neutrosophic weight of
one-to-one matching (i.e., M) is described by the following.

ω ( M) = ∑ ω ( E)
E∈ M

A matching (i.e., M) is called a perfect matching if each node of X is assigned to one of the Y
nodes. Therefore, a linear assignment problem is defined as determining a perfect matching in G that
is maximally weighted.

7.2. Road Transportation Network

A road transportation network can be modeled by a neutrosophic graph, due to its imprecise
dataset, neutrosophic information for vertices as the geometric location, the grade of significance,
and so on; in addition to the arcs as the width, length, quality, traffic, and so on. Therefore, we can
represent a road transportation network properly by using neutrosophic graphs where the vertices and
the edges represent the cities and routes between two cities, respectively. Dijkstra’s algorithm [45,46]
is a popular and well-known algorithm to solve the shortest path problem (SPP). It is an algorithmic
approach to find the shortest path between two cities within the road transportation network. Floyd
also contributed much to the efficient route-finding algorithms, hence the famous Floyd’s algorithm
to find the shortest paths in any network with weights. The SPP is very useful to model the network
routing protocol. In any shortest path algorithm, we consider a node as the source node and find all
the shortest paths between the source vertex and destination vertex using all other nodes in the graph.
Hence, the proposed dm -degree of the source node will pave a new way to find the shortest path.
When its efficiency is proven, it could be considered as the best one in the road transport network.

7.3. Social Networks

Social networks are platforms on which millions of people are interconnected everywhere in the
world. There are many social networks [47] like Facebook, WhatsApp, and Instagram. These networks
have reached a rapid progress in recent years and have spread all over the world in several ways.
There are also some security issues pointed out recently by ethical hacking teams; hence to maintain
the privacy of the user and the data released by them. In the medium, messages could be tampered
Mathematics 2019, 7, 551 18 of 20

with easily when a user sends the information to all the other users of his/her group. However, the
situation is that a user should be connected to all other users. Social networks can be modeled by
a neutrosophic regular graph, since dm -degree farthest vertices are assumed by paths; therefore, it
is a vertex that has at most dm -degrees; hence, the secrecy of this vertex is maintained, and the user
information is secured. If all the vertices of a social network have the same neutrosophic dm -degree,
then the user’s secrecy is maintained by adding a lowest arc and producing a dm -degree in these
graphs. Therefore, in these ways, in social networks, the data integrity is maintained, resisting more
neighborhood attacks and threats.
The concept of regularity plays an important role in modeling many real-life optimization
problems (e.g., social networks, wireless networks, road networks, assignment problems), so the main
focus of this study is to present the concept of regularity in neutrosophic graph theory. The degree
of a vertex in a graph is a way to find the number of relations of a vertex, so to analyze a network, it
is very essential to find the degree of vertices. In this study, we have described the different types of
degrees (dm -degree and tdm -degree) of nodes in a neutrosophic graph. The idea of the neutrosophic
degree of a node is very important from different aspects in the real-world. Here, we have presented
the definition of the free vertex (node) and busy vertex (node) of a regular neutrosophic graph, which
are very useful to find the solutions of many real-life problems like the traffic congestion problem.
Finally, we have mentioned some real-life applications of the regular neutrosophic graph, complete
bipartite neutrosophic graph, and node dm -degree of a neutrosophic graph. This study will be useful
when the neutrosophic graphs are extremely large. For example, we have described the utility of the
regular neutrosophic graph to model a social network. In the Twenty First Century, social networks
are platforms on which millions of people are interconnected everywhere in the world. Therefore,
as a future study, we will have to consider some large regular neutrosophic fuzzy graphs to model
the social networks in real-life scenarios, and the closeness and diameter need to be computed for
the analysis of this social network. Furthermore, we will try to propose some algorithmic methods to
determine the closeness and diameter of any social network in real life. Despite the demand for future
work, this paper is a useful initial contribution to neutrosophic graph theory, and some problems are
described under the neutrosophic environment.

8. Conclusions
The main contribution of this manuscript is to introduce the idea of regularity in neutrosophic
graph theory. In this paper, we have described the notion of the dm -degree and tdm -degree of nodes
in a neutrosophic graph. Some different types of neutrosophic graphs such as the regular, regular
strong, dm -regular, tdm -regular, and complete bipartite neutrosophic graph were introduced here.
We have also provided some sufficient criteria for which the tdm -regular neutrosophic graph and
dm -regular neutrosophic graph are equivalent. We have introduced the definition of the µ-complement
and h-morphism of a neutrosophic graph. Some properties of the complement and isomorphic regular
neutrosophic graph were also presented here. Finally, some real-life applications of the regular
neutrosophic graph and complete bipartite neutrosophic graph were described for the assignment
problem, road transportation network, and social networks. In the future, we will focus on the study of
neutrosophic intersection graphs, neutrosophic interval graphs, neutrosophic hyper graphs, and so on.
The idea of the neutrosophic graph can be used in several areas of expert systems, image processing,
computer networks, and social systems.

Author Contributions: L.H., A.D., Y.H., and Y.L. conceived of the presented idea. P.K.K.K., D.K., L.H., and A.D.
developed the theory and performed the computations. Y.H. and Y.L. verified the analytical methods. A.D.
contributed reagents/materials/analysis tools. L.H., Y.H., and Y.L. wrote the paper.
Funding: The study was supported by the “Major scientific and technological innovation project of Shandong
Province, China (Grant No. 2017CXGC0901)” and the ”Major basic research project of Shandong natural science
foundation, China (Grant No. ZR2018ZC0436)”.
Conflicts of Interest: The authors declare no conflict of interest.
Mathematics 2019, 7, 551 19 of 20

References
1. Zadeh, L.A. Fuzzy sets. Inf. Control 1965, 8, 338–353. [CrossRef]
2. Dey, A.; Pal, A.; Pal, T. Interval type 2 fuzzy set in fuzzy shortest path problem. Mathematics 2016, 4, 62.
[CrossRef]
3. Kaufmann, A. Introduction to the Theory of Fuzzy Subsets; Academic Press: Cambridge, MA, USA, 1975;
Volume 2.
4. Rosenfeld, A. Fuzzy graphs. In Fuzzy Sets and Their Applications to Cognitive and Decision Processes; Elsevier:
Amsterdam, The Netherlands, 1975; pp. 77–95.
5. Gani, A.N.; Radha, K. On regular fuzzy graphs. J. Phys. Sci. 2008, 12, 33–40.
6. Samanta, S.; Pal, M. Irregular bipolar fuzzy graphs. arXiv 2012, arXiv:1209.1682.
7. Rashmanlou, H.; Pal, M. Some properties of highly irregular interval-valued fuzzy graphs. World Appl. Sci. J.
2013, 27, 1756–1773.
8. Rashmanlou, H.; Samanta, S.; Pal, M.; Borzooei, R. A study on vague graphs. SpringerPlus 2016, 5, 1234.
[CrossRef] [PubMed]
9. Rashmanlou, H.; Samanta, S.; Pal, M.; Borzooei, R.A. A study on bipolar fuzzy graphs. J. Intell. Fuzzy Syst.
2015, 28, 571–580.
10. Pramanik, T.; Samanta, S.; Pal, M.; Mondal, S.; Sarkar, B. Interval-valued fuzzy φ-tolerance competition
graphs. SpringerPlus 2016, 5, 1981. [CrossRef]
11. Nandhini, S.; Nandhini, E. Strongly Irregular Fuzzy Graphs. Int. J. Math. Arch. EISSN 2229-5046 2014, 5,
2229–5046.
12. Ghorai, G.; Pal, M. Certain types of product bipolar fuzzy graphs. Int. J. Appl. Comput. Math. 2017,
3, 605–619. [CrossRef]
13. Dey, A.; Son, L.; Kumar, P.; Selvachandran, G.; Quek, S. New concepts on vertex and edge coloring of simple
vague graphs. Symmetry 2018, 10, 373. [CrossRef]
14. Borzooei, R.A.; Rashmanlou, H.; Samanta, S.; Pal, M. New concepts of vague competition graphs. J. Intell.
Fuzzy Syst. 2016, 31, 69–75. [CrossRef]
15. Borzooei, R.A.; Rashmanlou, H.; Samanta, S.; Pal, M. Regularity of vague graphs. J. Intell. Fuzzy Syst. 2016,
30, 3681–3689. [CrossRef]
16. Atanassov, K.T. Intuitionistic fuzzy sets. Fuzzy Sets Syst. 1986, 20, 87–96. [CrossRef]
17. Atanassov, K.T. Intuitionistic fuzzy sets. In Proceedings of the VII ITKR’s Session, Sofia, Bulgarian, 20–23
June 1983. (In Bulgarian)
18. Atanassov, K.T. Intuitionistic fuzzy sets. In Intuitionistic Fuzzy Sets; Springer: Berlin, Germany, 1999;
pp. 1–137.
19. Shannon, A.; Atanassov, K. On a generalization of intuitionistic fuzzy graphs. NIFS 2006, 12, 24–29.
20. Parvathi, R.; Karunambigai, M. Intuitionistic fuzzy graphs. In Computational Intelligence, Theory and
Applications; Springer: Berlin, Germany, 2006; pp. 139–150.
21. Parvathi, R.; Thamizhendhi, G. Domination in intuitionistic fuzzy graphs. Notes Intuit. Fuzzy Sets 2010,
16, 39–49.
22. Parvathi, R.; Karunambigai, M.; Atanassov, K.T. Operations on intuitionistic fuzzy graphs. In Proceedings
of the 2009 IEEE International Conference on Fuzzy Systems, Jeju Island, Korea, 20–24 August 2009;
pp. 1396–1401.
23. Rashmanlou, H.; Samanta, S.; Pal, M.; Borzooei, R.A. Intuitionistic fuzzy graphs with categorical properties.
Fuzzy Inf. Eng. 2015, 7, 317–334. [CrossRef]
24. Rashmanlou, H.; Samanta, S.; Pal, M.; Borzooei, R.A. Bipolar fuzzy graphs with categorical properties. Int. J.
Comput. Intell. Syst. 2015, 8, 808–818. [CrossRef]
25. Akram, M.; Akmal, R. Operations on intuitionistic fuzzy graph structures. Fuzzy Inf. Eng. 2016, 8, 389–410.
[CrossRef]
26. Rashmanlou, H.; Borzooei, R.; Samanta, S.; Pal, M. Properties of interval valued intuitionistic (S, T)—Fuzzy
graphs. Pac. Sci. Rev. A Nat. Sci. Eng. 2016, 18, 30–37. [CrossRef]
27. Sahoo, S.; Pal, M. Different types of products on intuitionistic fuzzy graphs. Pac. Sci. Rev. A Nat. Sci. Eng.
2015, 17, 87–96. [CrossRef]
Mathematics 2019, 7, 551 20 of 20

28. Smarandache, F. A Unifying Field in Logic: Neutrosophic Logic. Neutrosophy, Neutrosophic Set, Neutrosophic
Probability and Statistics; Infinite Study; Modern Science Publisher: New York, NY, USA, 2005.
29. Ye, J. Single-valued neutrosophic minimum spanning tree and its clustering method. J. Intell. Syst. 2014,
23, 311–324. [CrossRef]
30. Yang, H.L.; Guo, Z.L.; She, Y.; Liao, X. On single valued neutrosophic relations. J. Intell. Fuzzy Syst. 2016,
30, 1045–1056. [CrossRef]
31. Naz, S.; Rashmanlou, H.; Malik, M.A. Operations on single valued neutrosophic graphs with application.
J. Intell. Fuzzy Syst. 2017, 32, 2137–2151. [CrossRef]
32. Akram, M.; Waseem, N.; Dudek, W.A. Certain types of edge m-polar fuzzy graphs. Iran. J. Fuzzy Syst. 2017,
14, 27–50.
33. Akram, M. Certain bipolar neutrosophic competition graphs. J. Indones. Math. Soc. 2017, 24, 1–25.
34. Akram, M.; Siddique, S. Neutrosophic competition graphs with applications. J. Intell. Fuzzy Syst. 2017,
33, 921–935. [CrossRef]
35. Akrama, M. Single-Valued Neutrosophic Planar Graphs; Infinite Study; Modern Science Publisher: New York,
NY, USA, 2016.
36. Gani, A.N.; Latha, S. On irregular fuzzy graphs. Appl. Math. Sci. 2012, 6, 517–523.
37. Maheswari, N.S.; Sekar, C. Semi neighbourly irregular graphs. Int. J. Comb. Graph Theory Appl. 2015,
5, 135–144.
38. Darabian, E.; Borzooei, R.A.; Rashmanlou, H.; Azadi, M. New concepts of regular and (highly) irregular
vague graphs with applications. Fuzzy Inf. Eng. 2017, 9, 161–179. [CrossRef]
39. Dey, A.; Broumi, S.; Bakali, A.; Talea, M.; Smarandache, F. A new algorithm for finding minimum spanning
trees with undirected neutrosophic graphs. Granul. Comput. 2019, 4, 63–69. [CrossRef]
40. Broumi, S.; Bakali, A.; Talea, M.; Smarandache, F.; Dey, A. Spanning tree problem with Neutrosophic edge
weights. Proc. Comput. Sci. 2018, 127, 190–199. [CrossRef]
41. Broumi, S.; Talea, M.; Smarandache, F.; Bakali, A. Decision-making method based on the interval
valued neutrosophic graph. In Proceedings of the IEEE 2016 Future Technologies Conference (FTC),
San Francisco, CA, USA, 6–7 December 2016; pp. 44–50.
42. Broumi, S.; Talea, M.; Bakali, A.; Smarandache, F. On bipolar single valued neutrosophic graphs. J. New Theory
2016, 11, 84–102.
43. Karaaslan, F.; Davvaz, B. Properties of single-valued neutrosophic graphs. J. Intell. Fuzzy Syst. 2018,
34, 57–79. [CrossRef]
44. Broumi, S.; Smarandache, F.; Talea, M.; Bakali, A. Operations on Interval Valued Neutrosophic Graphs; Infinite
Study; Modern Science Publisher: New York, NY, USA, 2016.
45. Dijkstra, E.W. A note on two problems in connexion with graphs. Numer. Math. 1959, 1, 269–271. [CrossRef]
46. Cormen, T.H.; Leiserson, C.E.; Rivest, R.L.; Stein, C. Introduction to Algorithms, 2nd ed.; MIT Press: Cambridge,
MA, USA, 2001.
47. Zuo, C.; Pal, A.; Dey, A. New Concepts of Picture Fuzzy Graphs with Application. Mathematics 2019, 7, 470.
[CrossRef]

c 2019 by the authors. Licensee MDPI, Basel, Switzerland. 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/).