Applied Mathematical Sciences, Vol. 9, 2015, no.

59, 2901 - 2911

HIKARI Ltd, www.m-hikari.com
http://dx.doi.org/10.12988/ams.2015.46427

On JB-Semigroups
Joemar C. Endam and Jocelyn P. Vilela

Department of Mathematics and Statistics

College of Science and Mathematics
MSU-Iligan Institute of Technology
Tibanga, Iligan City, Philippines

Copyright c 2014 Joemar C. Endam and Jocelyn P. Vilela. This article is distributed
under the Creative Commons Attribution License, which permits unrestricted use, distribu-
tion, and reproduction in any medium, provided the original work is properly cited.

Abstract
In this paper, we introduce the notion of JB-semigroup. We prove
that every ring determines a JB-semigroup, but the converse need not
be true. We also introduce the notions of JB-field and JB-domain,
and we prove that every JB-field is a JB-domain and every finite JB-
domain is a JB-field. Moreover, we introduce the notion of JB-ideal of
JB-semigroup, and we construct quotient JB-semigroup via JB-ideal.
Furthermore, we introduce the notion of JB-homomorphism of JB-
semigroups, and we provide some of its properties.

Mathematics Subject Classification: 06F35, 03G25

Keywords: JB-semigroup, JB-field, JB-domain, JB-ideal, quotient JB-

semigroup, JB-homomorphism

1 Introduction
In 1966, Y. Imai and K. Iseki [3] introduced two classes of abstract algebras:
BCK-algebras and BCI-algebras. It is known that the class of BCK-algebras
is a proper subclass of the class of BCI-algebras. In 2002, J. Neggers and H.S.
Kim [7] introduced the notion of B-algebra. In 2006, K.H. Kim [4] introduced
the notion of KS-semigroup. He formulated this new algebra from a BCK-
algebra by combining the concept of semigroup, that is, adding an associative
2902 Joemar C. Endam and Jocelyn P. Vilela

2-ary operation to BCK-algebra. In view of this process, we introduce the

notion of JB-semigroup by adding the concept of semigroup to B-algebra.
In this paper, we prove that every ring determines a JB-semigroup, but the
converse need not be true. Furthermore, we introduce the notions of JB-
field and JB-domain, and we prove that every JB-field is a JB-domain and
every finite JB-domain is a JB-field. We also introduce the notion of JB-
ideal of JB-semigroup, and we construct quotient JB-semigroup via JB-ideal.
Furthermore, we introduce the notion of JB-homomorphism of JB-semigroups,
and we provide some of its properties.

2 JB-semigroups and Rings

Definition 2.1 [7] A B-algebra is an algebra (X; ∗, 0) of type (2, 0) satis-
fying the following axioms for all x, y, z ∈ X:

(I) x ∗ x = 0,

(II) x ∗ 0 = x,

(III) (x ∗ y) ∗ z = x ∗ (z ∗ (0 ∗ y)).

In [7], a B-algebra (X; ∗, 0) is called commutative if a ∗ (0 ∗ b) = b ∗ (0 ∗ a)

for all a, b ∈ X. A nonempty subset N of a B-algebra (X; ∗, 0) is called a
subalgebra of X if x ∗ y ∈ N for any x, y ∈ N . It is called normal in X if for
any x ∗ y, a ∗ b ∈ N implies (x ∗ a) ∗ (y ∗ b) ∈ N . A normal subset of X is a
subalgebra of X. Moreover, in [2], the intersection of any nonempty collection
of (normal) subalgebras of X is also a (normal) subalgebra of X.

Definition 2.2 A JB-semigroup is a nonempty set X together with two

binary operations ∗ and · and a constant 0 satisfying the following:

i. (X; ∗, 0) is a B-algebra;

ii. (X, ·) is a semigroup; and

iii. The operation · is left and right distributive over the operation ∗.

Definition 2.2(i) implies that if (X; ∗, ·, 0) is a JB-semigroup, then all prop-

erties pertaining to the binary operation ∗ with respect to the B-algebra
(X; ∗, 0) also hold for the JB-semigroup (X; ∗, ·, 0). In particular, the following
properties hold in a JB-semigroup: (P1) 0 ∗ (0 ∗ x) = x [7], (P2) x ∗ y = 0
implies x = y [7], and (P3) x ∗ y = 0 ∗ (y ∗ x) [8].

The following are examples of JB-semigroups.

On JB-semigroups 2903

Example 2.3 Let X = {0, a, b, c} be a set with the following tables:

∗ 0 a b c · 0 a b c
0 0 a b c 0 0 0 0 0
a a 0 c b a 0 a b c
b b c 0 a b 0 b c a
c c b a 0 c 0 c a b

Then by routine calculations, (X; ∗, ·, 0) is a JB-semigroup.

Example 2.4 Let X = {0, a, b, c} be a set with the following tables:

∗ 0 a b c · 0 a b c
0 0 0 0 0 0 0 0 0 0
a a 0 a a a 0 a 0 0
b b b 0 b b 0 0 b 0
c c c c 0 c 0 c 0 0

Then by routine calculations, (X; ∗, ·, 0) is a JB-semigroup.

Example 2.5 Let X = {0, a, b, c} be a set with the following tables:

∗ 0 a b c · 0 a b c
0 0 a b c 0 0 0 0 0
a a 0 c b a 0 0 a a
b b c 0 a b 0 0 b b
c c b a 0 c 0 0 c c

Then by routine calculations, (X; ∗, ·, 0) is a JB-semigroup.

Example 2.6 By routine calculations, (Z; ∗, ·, 0), (Q; ∗, ·, 0), (R; ∗, ·, 0), and
(C; ∗, ·, 0) are JB-semigroups, where x ∗ y = x − y and · is the usual multipli-
cation.

Lemma 2.7 Let (X; ∗, ·, 0) be a JB-semigroup and a, b, c ∈ X. Then the

following properties hold:

i. a · 0 = 0 · a = 0,

ii. a · (0 ∗ b) = (0 ∗ a) · b = 0 ∗ (a · b),

iii. (0 ∗ a) · (0 ∗ b) = a · b,

iv. a · (b ∗ (0 ∗ c)) = (a · b) ∗ (0 ∗ (a · c)), (b ∗ (0 ∗ c)) · a = (b · a) ∗ (0 ∗ (c · a)).

2904 Joemar C. Endam and Jocelyn P. Vilela

Proof : Let a, b, c ∈ X.
i. Now, a · 0 = a · (0 ∗ 0) = (a · 0) ∗ (a · 0) = 0. Similarly, 0 · a = 0.
ii. By (i), a · (0 ∗ b) = (a · 0) ∗ (a · b) = 0 ∗ (a · b) = (0 · b) ∗ (a · b) = (0 ∗ a) · b.
iii. By (i), (ii), and (P1), (0 ∗ a) · (0 ∗ b) = 0 ∗ (a · (0 ∗ b)) = 0 ∗ ((a · 0) ∗ (a · b)) =
0 ∗ (0 ∗ (a · b)) = a · b.
iv. By (ii), a · (b ∗ (0 ∗ c)) = (a · b) ∗ (a · (0 ∗ c)) = (a · b) ∗ (0 ∗ (a · c)). Similarly,
(b ∗ (0 ∗ c)) · a = (b · a) ∗ (0 ∗ (c · a)). 

To discuss the relationship of JB-semigroups and rings, we recall first the

relationship of B-algebras and groups. Now, every group determines a B-
algebra ([1], [7]), and if the given group is commutative, then the B-algebra is
commutative [1], as shown in the following theorem.
Theorem 2.8 If (X; ◦, e) is a group, then (X; ∗, 0 = e) is a B-algebra,
where x ∗ y = x ◦ y −1 . Moreover, if (X; ◦, e) is commutative, then (X; ∗, 0 = e)
is commutative.
This means that every group can be transformed into a B-algebra. It is
then a question of interest to determine whether or not all B-algebras can be
transformed into groups. The answer is affirmative, proved by M. Kondo and
Y.B. Jun [5], and if the given B-algebra is commutative, then the group is
commutative [1], as shown in the following theorem.
Theorem 2.9 If (X; ∗, 0) is a B-algebra, then (X; ◦, e = 0) is a group,
where x◦y = x∗(0∗y). Moreover, if (X; ∗, 0) is commutative, then (X; ◦, e = 0)
is commutative.
Combining Theorem 2.8 and Theorem 2.9, the class of B-algebras and the
class of groups coincide, in some sense. With regards to JB-semigroups and
rings, we prove that every ring determines a JB-semigroup.
Theorem 2.10 If (X; +, ·, e) is a ring, then (X; ∗, ·, 0 = e) is a JB-semigroup,
where x ∗ y = x − y.
Proof : Let (X; +, ·, e) be a ring. By Theorem 2.8, (X; ∗, 0 = e) is a B-
algebra. Clearly, (X, ·) is a semigroup since · is associative. If x, y, z ∈ X,
then x · (y ∗ z) = x · (y − z) = (x · y) − (x · z) = (x · y) ∗ (x · z). Similarly,
(x ∗ y) · z = (x · z) ∗ (y · z). Therefore, (X; ∗, ·, 0 = e) is a JB-semigroup. 

Let (X; ∗, ·, 0) be a JB-semigroup. If we define x + y = x ∗ (0 ∗ y), then

(X; +, ·, e = 0) need not be a ring. Consider the JB-semigroup in Example 2.4,
(X; +, ·, e = 0) is not a ring since a + b = a ∗ (0 ∗ b) = a 6= b = b ∗ (0 ∗ a) = b + a.
This means that there exists a JB-semigroup that cannot be transformed into
a ring, with respect to the predefined transformation. But if we restrict the
B-algebra (X; ∗, 0) to be commutative, then we get the following theorem.
On JB-semigroups 2905

Theorem 2.11 If (X; ∗, ·, 0) is a JB-semigroup such that (X; ∗, 0) is com-

mutative, then (X; +, ·, e = 0) is a ring, where x + y = x ∗ (0 ∗ y).

Proof : Let (X; ∗, ·, 0) be a JB-semigroup such that (X; ∗, 0) is commutative.

By Theorem 2.9, (X; +, e = 0) is a commutative group. Since (X, ·) is a
semigroup, · is associative. Let x, y, z ∈ X. Then by Lemma 2.7(ii), we have
x·(y +z) = x·(y ∗(0∗z)) = (x·y)∗(x·(0∗z)) = (x·y)∗(0∗(x·z)) = x·y +x·z.
Similarly, (x + y) · z = x · z + y · z. Therefore, (X; +, ·, e = 0) is a ring. 

3 JB-field and JB-domain

Throughout this section, X means a JB-semigroup (X; ∗, ·, 0).

Definition 3.1 A nonempty subset S of X is called a sub JB-semigroup of

X if x ∗ y, x · y ∈ S for all x, y ∈ S.

Clearly, 0 is an element of a sub JB-semigroup. Moreover, {0} and X

are sub JB-semigroups of X. The set S1 = {0, a} is a sub JB-semigroup of
the JB-semigroup in Example 2.3, while the set S2 = {0, a, b} is not since
a∗b=c∈ / S2 .

Definition 3.2 If a · b = b · a for all a, b ∈ X, then X is called commutative.

If X is not commutative, then it is called noncommutative.

Definition 3.2 means that X is commutative if and only if the semigroup

(X, ·) is commutative. The JB-semigroup in Example 2.3 is commutative,
while the JB-semigroup in Example 2.5 is noncommutative.

Definition 3.3 The subset C[X] = {a ∈ X: a · b = b · a for all b ∈ X} of

X is called the center of X.

Remark 3.4 X is commutative if and only if X = C[X].

Definition 3.5 An element y ∈ X is called a unity in X if x · y = x = y · x

for all x ∈ X.

The JB-semigroup in Example 2.3 has unity a, while JB-semigroup in Ex-

ampe 2.5 has no unity. The unity of a JB-semigroup is unique (if it exists),
and is denoted by 1.

Definition 3.6 Let 1 ∈ X. An element a ∈ X is called 1-invertible if there

exists b ∈ X such that a · b = 1 = b · a.
2906 Joemar C. Endam and Jocelyn P. Vilela

In Example 2.3, all elements of X not equal to 0 are 1-invertibles, that is,
elements a, b, and c are 1-invertibles.

Lemma 3.7 Let 1 ∈ X. Then the following statements hold:

i. (0 ∗ 1) · a = 0 ∗ a = a · (0 ∗ 1) for all a ∈ X,

ii. (0 ∗ 1) · (0 ∗ 1) = 1,

iii. if a is 1-invertible in X, then 0 ∗ a is 1-invertible in X.

Proof : Let a ∈ X. Then (0∗1)·a = (0·a)∗(1·a) = 0∗a. Similarly, a·(0∗1) = 0∗a.

This proves (i). By Lemma 2.7(iii), (0 ∗ 1) · (0 ∗ 1) = 1 · 1 = 1. This proves (ii).
Suppose a is 1-invertible in X. Then there exists b ∈ X such that a·b = 1 = b·a.
Now, 0 ∗ b ∈ X with (0 ∗ b) · (0 ∗ a) = b ∗ a = 1 = (a · b) = (0 ∗ a) · (0 ∗ b). Thus,
0 ∗ a is 1-invertible in X. This proves (iii). 

Theorem 3.8 Let 1 ∈ X and T be the set of all 1-invertible elements of

X. Then T 6= ∅, 0 ∈
/ T , and a · b ∈ T for all a, b ∈ T .

Proof : Let T be the set of all 1-invertible elements of X. Since 1 · 1 = 1 = 1 · 1,

1 ∈ T . Thus, T 6= ∅. Suppose that 0 ∈ T . Then there exists b ∈ X such
that 0 · b = 1 = b · 0. But 0 · b = 0 and so 0 = 1, a contradiction. Thus,
0∈ / T . Let a, b ∈ T . Then there exist c, d ∈ X such that a · c = 1 = c · a and
b · d = 1 = d · b. Now, (a · b) · (d · c) = a · (b · d) · c = a · (1) · c = a · c = 1 and
(d · c) · (a · b) = d · (c · a) · b = d · (1) · b = d · b = 1. Thus, a · b ∈ T . 

Definition 3.9 Let 1 ∈ X. Then X is called a JB-field if the semigroup

(X, ·) is commutative and every 0 6= a ∈ X is 1-invertible.

The JB-semigroup in Example 2.3 is a JB-field.

Definition 3.10 An element 0 6= a ∈ X is called 0-divisor if there exists

b ∈ X such that b 6= 0 and either a · b = 0 or b · a = 0.

Remark 3.11 An element cannot be a 1-invertible and 0-divisor at the

same time. Thus, a JB-field has no 0-divisors.

Theorem 3.12 If X has no 0-divisors, then left and right cancellation laws
hold, that is, for all a, b, c ∈ X, a 6= 0, a · b = a · c implies b = c (left
cancellation) and b · a = c · a implies b = c (right cancellation). If either left
or right cancellation law holds, then X has no 0-divisors.
On JB-semigroups 2907

Proof : Let a, b, c ∈ X such that a · b = a · c and a 6= 0. Then a · (b ∗ c) =

(a · b) ∗ (a · c) = 0. Since X has no 0-divisors and a 6= 0, we have b ∗ c = 0.
By (P2), b = c. Hence, the left cancellation law holds. Similarly, the right
cancellation law holds. Conversely, suppose one of the cancellation laws hold,
say, the left cancellation. Let a be a nonzero element of X and b ∈ X. Suppose
a · b = 0. Then a · b = a · 0 and so by left cancellation, b = 0. Suppose b · a = 0
and b 6= 0. Then b · a = b · 0 and so by left cancellation, a = 0, a contradiction.
Therefore, b = 0. Hence, X has no 0-divisors. Similarly, the right cancellation
law implies that X has no 0-divisors. 

Definition 3.13 Let 1 ∈ X. Then X is called a JB-domain if the semi-

group (X, ·) is commutative and X has no 0-divisors.

The JB-semigroup in Example 2.3 is a JB-domain.

Remark 3.14 Every JB-field is a JB-domain.

The converse of Remark 3.14 need not be true. The JB-semigroup (Z; ∗, ·, 0)
in Example 2.6 is a JB-domain, but not a JB-field.

Theorem 3.15 A finite commutative JB-semigroup X with more than one

element and without 0-divisors is a JB-field.

Proof : Let a1 , a2 , . . . , an be the distinct elements of X. Let a ∈ X with a 6= 0.

Now, a · ai ∈ X for all i = 1, 2, . . . , n and so {a · a1 , a · a2 , . . . , a · an } ⊆
X. If a · ai = a · aj , then by Theorem 3.12, ai = aj . Thus, the elements
a · a1 , a · a2 , . . . , a · an are distinct and so X = {a · a1 , a · a2 , . . . , a · an }. Hence,
one of the elements must be equal to a, say, a·ai = a. Since X is commutative,
ai ·a = a·ai = a. Let b ∈ X. Then there exists aj ∈ X such that b = a·aj . Thus,
b·ai = ai ·b = ai ·(a·aj ) = (ai ·a)·aj = a·aj = b. This implies that ai is the unity
of X. We denote the unity of X by 1. Now, 1 ∈ X = {a·a1 , a·a2 , . . . , a·an } and
so one of the elements must be equal to 1, say, a · aj = 1. By commutativity,
aj ·a = a·aj = 1. Hence, every nonzero element of X is 1-invertible. Therefore,
X is a JB-field. 

Corollary 3.16 Every finite JB-domain is a JB-field.

4 JB-ideals and JB-homomorphisms

Definition 4.1 A nonempty subset I of X is called a JB-ideal of X if the
following hold:

i. (x ∗ a) ∗ (y ∗ b) ∈ I for any x ∗ y, a ∗ b ∈ I,
2908 Joemar C. Endam and Jocelyn P. Vilela

ii. a · x, x · a ∈ I for any a ∈ I, x ∈ X.

Definition 4.1(i) means that I is a normal subalgebra of the B-algebra

(X; ∗, 0). Obviously, the subsets {0} and X are JB-ideals of X. These JB-
ideals are called trivial JB-ideals. All other JB-ideals are called nontrivial
JB-ideals. Consider the JB-semigroup X in Example 2.4, the set I = {0, b} is
a JB-ideal of X, while the set J = {0, a} is not since c · a = c ∈
/ J.

Remark 4.2 Let I be a JB-ideal of X. Then I is a sub JB-semigroup of

X and I is a JB-ideal for every sub JB-semigroup of X containing I.

\4.3 Let {Iα : α ∈ A } be a nonempty collection of JB-ideals of

Theorem
X. Then Iα is a JB-ideal of X.
α∈A

Let ∅ 6= A, B ⊆ X. Define A ∗ B as the set {a ∗ (0 ∗ b): a ∈ A, b ∈ B}.

In [2], if A is a subalgebra of a B-algebra (X; ∗, 0), then A ∗ A = A.

Moreover, if A and B are normal subalgebras of X, then A ∗ B = B ∗ A is a
normal subalgebra of X.

Theorem 4.4 Let A, B, and C be JB-ideals of X. Then the following

properties hold:

i. A ∗ B = B ∗ A is a JB-ideal of X,

ii. A ∗ A = A,

iii. (A ∗ B) ∗ C = A ∗ (B ∗ C).

Proof : i. Let y ∈ A ∗ B and x ∈ X. Then y = a ∗ (0 ∗ b) for some a ∈ A, b ∈ B.

Since A and B are JB-ideals, a · x ∈ A and b · x ∈ B. Hence, by Lemma 2.7(ii),
we have y · x = (a ∗ (0 ∗ b)) · x = (a · x) ∗ ((0 ∗ b) · x) = (a · x) ∗ (0 ∗ (b · x)) ∈ A ∗ B.
Similarly, x · y ∈ A ∗ B. Therefore, A ∗ B = B ∗ A is a JB-ideal of X.
iii. By (III), (P1), and (P3), we have x ∈ (A∗B)∗C ⇔ x = (a∗(0∗b))∗(0∗c) ⇔
x = a ∗ [(0 ∗ c) ∗ (0 ∗ (0 ∗ b))] ⇔ x = a ∗ ((0 ∗ c) ∗ b) ⇔ x = a ∗ [0 ∗ (b ∗ (0 ∗ c))] ⇔
x ∈ A ∗ (B ∗ C). Hence, (A ∗ B) ∗ C = A ∗ (B ∗ C). 

In [7], if N is a normal subalgebra of a B-algebra (X; ∗, 0), then we have

a B-algebra (X/N ; ∗, [0]N ), where X/N = {[x]N : x ∈ X} and ∗ is defined by
[x]N ∗ [y]N = [x ∗ y]N . For x ∈ X, [x]N is the equivalence class containing
x, that is, [x]N = {y ∈ X: x ∼N y}, where x ∼N y if and only if x ∗ y ∈ N
for any x, y ∈ X. The algebra X/N is called the quotient B-algebra of X by N .
On JB-semigroups 2909

Let I be a JB-ideal of X. Then I is a normal subalgebra of the B-algebra

(X; ∗, 0) and (X/I; ∗, [0]I ) is a B-algebra. Now, define · on X/I by [x]I ·
[y]I = [x · y]I . The binary operation · on X/I is well-defined. To see this,
let [x]I = [x0 ]I and [y]I = [y 0 ]I . Then x ∗ x0 ∈ I and y ∗ y 0 ∈ I. Since I is
a JB-ideal, (x · y) ∗ (x · y 0 ) = x · (y ∗ y 0 ) ∈ I and so x · y ∼I x · y 0 . Also,
(x · y 0 ) ∗ (x0 · y 0 ) = (x ∗ x0 ) · y 0 ∈ I and so x · y 0 ∼I x0 · y 0 . Thus, x · y ∼I x0 · y 0 .
Hence, [x]I · [y]I = [x · y]I = [x0 · y 0 ]I = [x0 ]I · [y 0 ]I .

Theorem 4.5 Let I be a JB-ideal of X. Then (X/I; ∗, ·, [0]I ) is a JB-

semigroup, where ∗ and · defined as above. If X is commutative or has a
unity, then the same is true of X/I.

The JB-semigroup X/I in Theorem 4.5 is called quotient JB-semigroup of

X by I.

Let (X; ∗, ·, 0X ) and (Y ; ∗, ·, 0Y ) be JB-semigroups. A map ϕ : X → Y

is called a JB-homomorphism from X into Y if ϕ(x ∗ y) = ϕ(x) ∗ ϕ(y)
and ϕ(x · y) = ϕ(x) · ϕ(y) for any x, y ∈ X. A JB-homomorphism ϕ is
called a JB-monomorphism, JB-epimorphism, or JB-isomorphism if ϕ is one-
to-one, onto, or a bijection, respectively. A JB-homomorphism ϕ : X → X
is called a JB-endomorphism and a JB-isomorphism ϕ : X → X is called a
JB-automorphism.

Remark 4.6 Let ϕ : X → Y be a JB-homomorphism. Then ϕ maps 0X to

0Y , that is, ϕ(0X ) = 0Y , and for all x ∈ X, ϕ(0X ∗ x) = 0Y ∗ ϕ(x).

Lemma 4.7 If ϕ: X → Y and ψ: Y → Z are JB-homomorphisms, then

ψ ◦ ϕ: X → Z is also a JB-homomorphism.

Assuming compatibility of functions so that composition is defined, the

following corollary easily follows.

Corollary 4.8 The composition of JB-monomorphisms is a JB-monomorphism,

the composition of JB-epimorphisms is a JB-epimorphism, the composition of
JB-isomorphisms is a JB-isomorphism, and the composition of JB-automorphisms
is a JB-automorphism.

Let ϕ : X → Y be a B-homomorphism, that is, ϕ(x ∗ y) = ϕ(x) ∗ ϕ(y).

In [2], if J is a normal subalgebra of Y , then ϕ−1 (J) is a normal subalgebra
of X. If I is a normal subalgebra of X and ϕ is onto, then ϕ(I) is a normal
subalgebra of Y .

Lemma 4.9 Let ϕ: X → Y be a JB-homomorphism.

2910 Joemar C. Endam and Jocelyn P. Vilela

i. If I is a sub JB-semigroup of X, then ϕ(I) is a sub JB-semigroup of Y .

Moreover, if I is commutative, then ϕ(I) is commutative.
ii. If J is a sub JB-semigroup of Y , then ϕ−1 (J) is a sub JB-semigroup of
X containing Ker ϕ.
iii. If I is a JB-ideal X and ϕ is onto, then ϕ(I) is a JB-ideal of Y .
iv. If J is a JB-ideal of Y , then ϕ−1 (J) is a JB-ideal of X.
Proof : Let ϕ: X → Y be a JB-homomorphism.
iii. Let a ∈ ϕ(I) and y ∈ Y . Since ϕ be onto, there exists b ∈ I such that
ϕ(b) = a. Since ϕ is onto, there exists x ∈ X such that ϕ(x) = y. Since I is a
JB-ideal of X, we have b · x, x · b ∈ I. Thus, a · y = ϕ(b) · ϕ(x) = ϕ(b · x) ∈ ϕ(I)
and y · a = ϕ(x) · ϕ(b) = ϕ(x · b) ∈ ϕ(I). Therefore, ϕ(I) is a JB-ideal of Y .
iv. Let a ∈ ϕ−1 (J) and x ∈ X. Then ϕ(a) ∈ J and ϕ(x) ∈ Y . Since J is
a JB-ideal of Y , ϕ(a · x) = ϕ(a) · ϕ(x) ∈ J and ϕ(x · a) = ϕ(x) · ϕ(a) ∈ J.
Therefore, a · x, x · a ∈ ϕ−1 (J) and so ϕ−1 (J) is a JB-ideal of X. 
Definition 4.10 Let ϕ: X → Y be a JB-homomorphism from (X; ∗, ·, 0X )
into (Y ; ∗, ·, 0Y ). We define Ker ϕ to be the set Ker ϕ = {x ∈ X: ϕ(x) = 0Y }.
In [6], if ϕ : X → Y is a B-homomorphism, then Ker ϕ is a normal
subalgebra of X, and ϕ is one-to-one if and only if Ker ϕ = {0X }.
Lemma 4.11 Let ϕ : X → Y be a JB-homomorphism. Then
i. ϕ is one-to-one if and only if Ker ϕ = {0X }.
ii. Ker ϕ is a JB-ideal of X.
Proof : ii. Let a ∈ Ker ϕ and x ∈ X. Then ϕ(a) = 0Y . Hence, by Lemma 2.7(i),
ϕ(a·x) = ϕ(a)·ϕ(x) = 0Y ·ϕ(x) = 0Y and ϕ(x·a) = ϕ(x)·ϕ(a) = ϕ(x)·0Y = 0Y .
Therefore, a · x, x · a ∈ Ker ϕ and so Ker ϕ is a JB-ideal of X. 
Proposition 4.12 Let I be a JB-ideal of X. Then the map γ : X → X/I,
given by γ(x) = [x]I , is a JB-epimorphism, and Ker γ = I.
Proof : For all x, y ∈ X, γ(x ∗ y) = [x ∗ y]I = [x]I ∗ [y]I = γ(x) ∗ γ(y) and
γ(x · y) = [x · y]I = [x]I · [y]I = γ(x) · γ(y). Hence, γ is a JB-homomorphism.
Since γ is obviously onto, γ is a JB-epimorphism. Furthermore, Ker γ =
{x ∈ X: [x]I = I} = {x ∈ X: x ∈ I} = I. 

The mapping γ in Proposition 4.12 is called the natural (or canonical ) JB-
homomorphism of X onto X/I.

Acknowledgements. This research is funded by the Department of Sci-

ence and Technology (DOST) through Accelerated Science and Technology
Human Resource Development Program (ASTHRDP).
On JB-semigroups 2911

References
[1] P. J. Allen, J. Neggers, and H. S. Kim, B-algebras and groups, Scientiae
Mathematicae Japonicae Online, 9(2003), 159-165.

[2] J. C. Endam and J. P. Vilela, The Second Isomorphism Theorem

for B-algebras, Applied Mathematical Sciences, 8(38)(2014), 1865-1872.
http://dx.doi.org/10.12988/ams.2014.4291

[3] Y. Imai and K. Iseki, On axiom systems of proposi-

tional calculi XIV, Proc. Japan Academy, 42(1966), 19-22.
http://dx.doi.org/10.3792/pja/1195522169

[4] K. H. Kim, On Structure of KS-semigroups, International Mathematical

Forum, 1(2)(2006), 67-76.

[5] M. Kondo and Y. B. Jun, The Class of B-algebras Coincides with the Class
of Groups, Scientiae Mathematicae Japonicae Online, 7(2002), 175-177.

[6] J. Neggers and H. S. Kim, A fundamental theorem of B-homomorphism

for B-algebras, Int. Math. J., 2(3)(2002), 207-214.

[7] J. Neggers and H. S. Kim, On B-algebras, Mat. Vesnik, 54(2002), 21-29.

[8] A. Walendziak, Some Axiomatizations of B-algebras, Math. Slovaca,

56(3)(2006), 301-306.

[9] A. Walendziak, A note on normal subalgebras in B-algebras, Scientiae

Mathematicae Japonicae Online, (2005), 49-53.

Received: June 11, 2014; Published: April 12, 2015