International Mathematical Forum, 1, 2006, no.

23, 1113 - 1124

On fuzzy topological K -algebras

M. Akram University College of Information Technology University of the Punjab, Old Campus, Lahore-54000, Pakistan m.akram@pucit.edu.pk K. H. Dar Govt. College University Lahore, Department of Mathematics Katchery Road, Lahore-54000, Pakistan prof khdar@yahoo.com Abstract. In this paper, we introduce the notions of fuzzy topological subalgebras and ideals in K -algebras, and investigate some of their properties. We also discuss the properties of homomorphic image and inverse image of fuzzy topological ideals of K -algebras. Keywords: (Fuzzy )continuous maps, (Fuzzy topological )ideals, Hausdor spaces ,C5 -disconnectness, Compactness. Mathematics Subject Classication: 06F35, 94D05.

Introduction

K. H. Dar and M. Akram [9] introduced a class of logical algebras: K -algebras on a group G( briey, K (G)-algebras) using the induced binary operation and have further extended its scope of study in [10] . Fuzzy set was introduced by Zadeh [4]. Since then there have been wide-ranging applications of the fuzzy set theory. Many research workers have fuzzied the various mathematical structures, such as topological spaces, functional analysis, loop, group, ring, near ring, vector spaces, automation. In this paper, we introduce the notions of fuzzy topological subalgebras and ideals in K -algebras, and investigate some of their properties such as fuzzy Hausdor spaces, fuzzy C5 -disconnectness and fuzzy compactness. We also discuss the properties of homomorphic image and inverse image of fuzzy topological ideals of K -algebras.

Preliminaries

In this section, we review some denitions and properties that will be used in the sequel:

1114

M. Akram and K. H. Dar

Denition 2.1. [9] Let (G, , e) be a group with the identity e such that x2 = e for some x(= e) G. A K -algebra is a structure K = (G, , , e), where is a binary operation on G which is induced from the operation , that satises the following: (k1) (a, x, y G) ((a (k2) (a, x G) (a (k3) (a G) (a (k4) (a G) (a (k5) (a G) (e (a a = e), e = a), a = a1 ). x) (a y ) = (a x1 ) (y 1 a), x1 )) a),

x) = (a

If G is abelian , then conditions (k1) and (k2) can be written as follows: (k1 ) (a, x, y G) ((a (k2 ) (a, x G) (a respectively. Denition 2.2. [9] A nonempty subset H of a K -algebra K is called a subalgebra of K if it satises: (i) e H, (ii) (a, b H ) (a b H ). (a x) (a y) = y x),

x) = x),

Denition 2.3. Let K1 = (G1 , , , e1 ) and K2 = (G2 , , , e2 ) be K -algebras. A mapping f : K1 K2 of K -algebras is called a homomorphism if f (x y ) = f (x) f (y ), x, y K1 . Note that if f is a homomorphism, then f (e1 ) = e2 . Denition 2.4. A nonempty subset A of a K -algebra K is called an ideal of K if, (i) e A, (ii) (y K)(x A) (x y ) A and y (y x) A x A).

Denition 2.5. [1] Let X be a non-empty set. A fuzzy (sub)set A in X is characterized by a membership function A : X [0, 1]. Denition 2.6. [1] Let A and B be two fuzzy sets in X , then following operations are valid :

On fuzzy topological K -algebras

1115

(a) (x X )(A B A (x) B (x)), (b) (x X )(A = B A (x) = B (x)) , (c) (x X )(A B = min (A (x), B (x))), (d) (x X )(A B = max (A (x), B (x))). Denition 2.7. The fuzzy sets X and 1X in X are dened by X = {x X : (x) = 0} and 1X = {x X : (x) = 1} respectively. Denition 2.8. A fuzzy topology on a set X is a family of fuzzy sets in X which satises the following conditions: (i) X , 1X , (ii) If A1 , A2 , then A1 A2 , (iii) If Ai for all i I , then iI Ai . The pair (X, ) is called a fuzzy topological space(FTS) and members of are called open fuzzy sets( OFSs), and the complement of fuzzy open sets are closed fuzzy sets(CFSs). Denition 2.9. [5] Let f be a mapping from a set X into set Y . (a) Let B be a fuzzy set in Y with membership function B . The inverse image of B , denoted by f 1 (B ), is the fuzzy set in X with membership for all x X . function f 1 (B) dened by f 1 (B) (x) = B (f (x)) (b) Let A be a fuzzy set in X with membership function A . The image of A, is denoted by f (A), is the fuzzy set in Y with membership function f (A) such that fsup (A )(y ) = supxf 1 (y) A (x), if f 1 (y ) = , 0, otherwise .

Proposition 2.10. Let A (Ai )be a fuzzy set with membership function A (Ai ) in X and B be a fuzzy set with membership function B in Y . Let f : X Y be a function. Then (a) If f is surjective , then f (f 1(B )) = B . (b) f (X ) = Y . (c) f (1X ) = 1Y , if f is surjective. (d) f 1 (1Y ) = 1X . (e) f 1 (Y ) = X . (f) f (Ai ) = f (Ai ).

1116

M. Akram and K. H. Dar

Fuzzy topological subalgebras

Denition 3.1. A fuzzy topology on a K -algebra K is said to be an indiscrete fuzzy topology if its only elements are empty fuzzy set (K ) and whole fuzzy set (1K ). A fuzzy topology on a K -algebra K is said to be a discrete fuzzy topology if it contains all fuzzy subsets of K. Denition 3.2. [?] A fuzzy set A in a K -algebra K with membership function A is called a fuzzy subalgebra of K if A (x y ) min{A (x), A (y )}, for all x, y K . Example 3.3. [9] Consider the K -algebra K = (G, , , e), where is given by the G = {e, a, a2 , a3 , a4 } is the cyclic group of order 5 and following Cayley table: e a a2 a3 a4 e e a a2 a3 a4 a a4 e a a2 a3 a2 a3 a4 e a a2 a3 a2 a3 a4 e a a4 a a2 e3 e4 e

(a) If we dene a fuzzy set as follows: A =< x, ( e a a2 a3 a4 , , , , )> 0.5 0.4 0.3 0.3 0.2 e a a2 a3 a4 , , , , )> 0.6 0.4 0.5 0.5 0.3

B =< x, (

Then the family {K , 1K , A, B } of fuzzy sets in K is a fuzzy topology on K because the empty fuzzy set K and the whole fuzzy set 1K are in , and the intersection of any two members of is a member of , and arbitrary union of members of is a member of . (b) Dene a fuzzy set A in K with membership function A dened by A (e) = 0.8 and A (x) = 0.02 for all x = e in K. It is easy to check that A is a fuzzy subalgebra of K.
) be fuzzy subspaces of fuzzy topoDenition 3.4. Let (F1 , F1 ) and (F2 , F 2 logical spaces (K1 , ) and (K2 , ) respectively, and let f be a mapping from ) if f (F1 ) F2 . (K1 , ) to (K2 , ). Then f is a mapping of (F1 , F1 ) into (F2 , F 2 Furthermore, f is relatively fuzzy continuous if for each open fuzzy set VF2 in F2 , the intersection f 1 (VF2 ) F1 is in F1 . Moreover, f is relatively fuzzy open if for each open fuzzy set UF1 in F1 , the image f (UF1 ) is in .

On fuzzy topological K -algebras

1117

Theorem 3.5. Let K1 and K2 be K -algebras and let (F, F ) and (G, G ) be fuzzy subspaces of (K1 , ) and (K2 , ) respectively. Let f be a fuzzy continuous mapping of K1 into K2 such that f (F1 ) F2 . Then f is relatively fuzzy continuous mapping of F1 into F2 .
, then there exists V such that Proof. Let VF2 be a fuzzy set in F 2 VF2 = V F2 . Since f is fuzzy continuous, it follows that f 1 (V ) is a fuzzy set in . Hence f 1 (VF2 ) F1 = f 1 (V F2 ) F1 = f 1 (V ) f 1(F2 ) F1 =f 1 (V ) F1 is a fuzzy set in F1 . This completes the proof.

Denition 3.6. Let 1 and 2 be fuzzy topologies on K -algebras K1 and K2 respectively and A be a fuzzy set with membership function A . A function f : (K1 , 1 ) (K2 , 2 ) is said to be a fuzzy continuous map from (K1 , 1 ) to (K2 , 2 ) if it satises following conditions: (i) For every A 2 , f 1 (A) 1 , (ii) For every fuzzy subalgebras A(of K2 ) in 2 , f 1 (A) is a fuzzy subalgebra (of K1 ) in 1 . Theorem 3.7. If 1 is a fuzzy topology on the K -algebra K1 and 2 is an indiscrete fuzzy topology on the K -algebras K2 , then every function f : (K1 , 1 ) (K2 , 2 ) is a fuzzy continuous map. Proof. Since 2 is an indiscrete fuzzy topology, therefore, 2 = {K2 , 1K2 }. Let f : (K1 , 1 ) (K2 , 2 ) be any mapping of K -algebras. We see that every member of 2 is a fuzzy subalgebra of K -algebra K2 . So it is enough to prove that for every A 2 , f 1 (A) 1 . Let K2 2 , then for any x K1 , f 1 (K2 )(x) = K2 (f (x)) = 0 [as f (x) K2 ] = K1 (x) [by def inition of empty f uzzy set]. Thus (f 1 (K2 )) = K1 1 . On the other hand, if 1K2 2 and x K1 , then we have (f 1 (1K2 ))(x) = 1K2 (f (x)) = 1 [as f (x) K2 ] = 1K1 (x) [by def inition of whole f uzzy set]. Thus (f 1 (1K2 )) = 1K1 1 . Hence f is a fuzzy continuous map which maps (K1 , 1 ) into (K2 , 2 ).

1118

M. Akram and K. H. Dar

Theorem 3.8. Let 1 and 2 be any two discrete fuzzy topologies on the K algebras K1 and K2 respectively. Then every homomorphism f : (K1 , 1 ) (K2 , 2 ) is a fuzzy continuous map. Proof. Since 1 and 2 are discrete fuzzy topologies on K -algebras K1 and K2 respectively, for every A 2 , f 1 (A) 1 [Note that f is not the usual inverse homomorphism from K2 to K1 ]. Let A be a fuzzy subalgebra in 2 with membership function A , then for any x, y K1 , we have (f 1 (A ))(x = A (f (x) y ) = A (f (x y ))

f (y )) [since f is homomorphism]

min{A (f (x)), A (f (y ))} [f or A is a f uzzy subalgebra of K2 ] = min{(f 1 (A ))(x), (f 1 (A )(y )}. This shows that f 1 (A) is a fuzzy subalgebra (of K1 ) in 1 and hence f is a fuzzy continuous map from K1 to K2 . Theorem 3.9. Let 1 and 2 be two fuzzy topologies dened on the K -algebras K1 and K2 , respectively. Then every homomorphism f : (K1 , 1 ) (K2 , 2 ) need not in general be a fuzzy continuous map. Proof. To prove this theorem it is sucient if we prove the result to be false for a particular 1 and 2 dened on any K -algebra K as in our denition of a fuzzy continuous map we have not assumed K1 and K2 to be distinct. Let K be any K -algebra. Dene two fuzzy topologies 1 and 2 on K -algebra K as 1 = {0K , 1K , } and 2 = {0K , 1K , }, where , : K [0, 1] dened as follows: 1 if x = e (x) = 0 x=e and (x) = 1 if x = e 0 x=e

where e is identity of a K -algebra. Dene f : (K, 1 ) (K, 2 ) by f (x) = x for all x K. Clearly, f is a homomorphism. For x K and 2 , we have (f 1 ())(x) = (f (x)) = (x) This gives (f 1 ())(x) = (x), for all x K. That is f 1 () = . Thus f 1 () / 1 , as / 1 . Hence f is not a fuzzy continuous map on K.

On fuzzy topological K -algebras

1119

Denition 3.10. Let (K1 , 1 ) and (K2 , 2 ) be any two fuzzy topological spaces. A function f : (K1 , 1 ) (K2 , 2 ) is said to be a fuzzy homomorphism if it satises the following conditions: f is bijective, both f and f 1 are fuzzy continuous maps. Denition 3.11. A fuzzy topological space (K, ) is said to be a fuzzy Hausdor space if and only if for any two distinct fuzzy points x and y there exist open fuzzy sets F and G such that F G = K . Theorem 3.12. Let 1 and 2 be the topologies on K -algebras K1 and K2 respectively and let f : K1 K2 be a fuzzy homeomorphism. Then K1 is a fuzzy Hausdor space if and only if K2 is a fuzzy Hausdor space. Proof. Suppose that K1 is a fuzzy Hausdor space. Let xt and xs be the fuzzy points in 2 with x = y (x, y K1 ), then f 1 (x) = f 1 (y ), as f is one to one. For z X , we consider (f 1 (x1 ))(z ) = x1 (f (z )) = = t (0, 1], if f(z)=x; 0, if f(z)= x. t (0, 1], if z = f 1 (x); 0, if z = f 1 (x).

= (f 1 (x))1 (z ). That is, (f 1 (xt ))(z ) = (f 1 (x))t (z ), for all z K1 . Hence, we have f 1 (xt ) = (f 1 (x))t . Similarly we can prove that f 1 (xs ) = (f 1 (x))s . By denition of a fuzzy Hausdor space, there exist open fuzzy sets Fx and Gy of f 1 (xt ) and f 1 (xs ) respectively such that Fx Gy = K1 . Since f is a fuzzy continuous map from K1 to K2 and f 1 is a fuzzy continuous map from K2 to K1 , there exist open fuzzy sets f (Fx ) and f (Gy ) of xt and ys respectively such that f (Fx ) f (Gy ) = f (Fx Gy ) = f (K1 ) = K2 . Hence K2 is a fuzzy Hausdor space. Conversely, let (K2 , 2 ) be a fuzzy Hausdor space. By a similar argument and by also using the fact that both f and f 1 are fuzzy continuous maps we can prove that (K1 , 1 ) is a fuzzy Hausdor space. The proof is now completed. Denition 3.13. Let be a fuzzy topology on a K -algebra K. A FTS (K, ) is said to be a fuzzy C5 - disconnected if there exists a fuzzy open and fuzzy closed set F with membership F such that F = 1K and F = K . A FTS (K, ) is said to be a fuzzy C5 - connected if it is not a fuzzy C5 disconnected.

1120

M. Akram and K. H. Dar

Theorem 3.14. Let 1 and 2 be the fuzzy topologies on K (G)-algebras K1 and K2 respectively. Let f : K1 K2 be a fuzzy continuous surjective mapping. If (K1 , 1 ) is a fuzzy C5 -connected, then (K2 , 2 ) is a fuzzy C5 -connected. Proof. Assume that (K2 , 2 ) is a fuzzy C5 -disconnected. Then there exist a fuzzy open and closed set F with membership function F such that F = 1K2 and F = K2 . Since f is a fuzzy continuous mapping, f 1 (F ) is both OFS and CFS. Thus f 1 (F )= 1K1 or f 1 (F )= K1 which is impossible.[since F = f (f 1 (F )) = f (1K1 ) = 1K2 , and F = f (f 1 (F )) = f (K1 ) = K2 .] This is contradiction to our assumption. Hence (K2 , 2 ) is also a fuzzy C5 connected. Denition 3.15. Let be a fuzzy topology on a K -algebra K and F be a fuzzy set in K with membership function F . If a class {< x, Fi } >: i I } of OFS in K satises the condition F {< x, Fi >: i I } , then it is called a fuzzy open cover of F . A nite subclass of the fuzzy open cover {< x, Fi >: i I } of F , which is also a fuzzy open cover of F , is called a nite subcover of {< x, Fi >: i I }. A FS F =< x, F > in a FTS (X, ) is called a fuzzy compact, if every fuzzy open cover of F has a nite subcover. Theorem 3.16. Let 1 and 2 be the fuzzy topologies on K -algebras K1 and K2 respectively. Let f : K1 K2 be a fuzzy continuous mapping. If F is a fuzzy compact in (K1 , 1 ), then f (F ) is a fuzzy compact in (K2 , 2 ). Proof. Let A= {Fi : i I }, where Fi =< y, Fi > be a fuzzy open cover of f (F ). Then B= {f 1(Fi ) : i I } is a fuzzy open cover of F . Since F is a fuzzy compact, there exists an nite subcover Fi (i = 1, 2, , n) of F such 1 that F n (Fi ). Thus i=1 f
1 (Fi )) f (F ) f (n i=1 f 1 (Fi )) f (F ) n i=1 f (f

f (F ) n i=1 Fi follows. Hence f (F ) is a fuzzy compact in (K2 , 2 ). Now we obtain a Theorem of relative fuzzy homomorphism inspired by [5]. Theorem 3.17. Let K be a K -algebra and let be a fuzzy topology on K algebra K. Let A be a fuzzy topological algebra in K. Then the inversion map f : A A dened by f (x) = x1 and the inner automorphism h : A A dened by h(g ) = aga1 are all relative fuzzy homomorphisms, where a {x : A(x) = A(e)}.

On fuzzy topological K -algebras

1121

Proof. Clearly f is one-to-one. Since f (A)(y ) =

z f 1 ( y )

sup A(z ) = A(y 1) = A(y ) y A.

That is, f (A) = A. Since f 1 (x) = x1 is relatively fuzzy continuous, f is relatively fuzzy open. Thus f is a relative fuzzy homomorphism. Let ra : A A be a right translation dened by ra (x) = xa and la : A A be left translation dened by la (x) = ax. Then (ra (A))(x) = = = sup A(z ) = A(xa1 )
1 z r a ( x)

min(A(x), A(a1 )) = min(A(x), A(e)) A(x) = A(xa1 a) min(A(xa1 ), A(a)) A(xa1 ) = (ra (A))(x).

Thus ra (A) = A. Let : A A A be a map dened by (x) = (x, a) and : A A A be a map dened by (x, y ) = xy . Then ra = . 1 = r a1 , Since and are fuzzy continuous, ra is fuzzy continuous. Since ra ra is a fuzzy homomorphism. Similarly, la is a fuzzy homomorphism. Since h is a composition of ra1 and la , h is relatively fuzzy homomorphism. This completes the proof.

Fuzzy topological ideals

Denition 4.1. A fuzzy set A in a K -algebra K with membership function A is called a fuzzy ideals of K if it satises: (i) (x G) (A (e) A (x)), (ii) (x, y G) (A (x) min{A (x y, A(y (y x))}).

Example 4.2. [9] Consider the K (S3 )-algebra K = (S3 , , , e) on the symmetric group S3 = {e, a, b, x, y, z } where e = (1), a = (123), b = (132), x = (12), y = (13), z = (23), and is given by the following Cayley table: e x y z a b e e x y z a b x x e b a z y y y a e b x z z z b a e y x a b z x y e a b a y z x b e

1122

M. Akram and K. H. Dar

Dene a fuzzy set A in K with membership function A by A (e) = t1 , A (p) = t2 , for all p = e, where t1 > t2 in [0, 1]. It is easy to check that A is a fuzzy ideal of K. Denition 4.3. Let K be a K -algebra and a fuzzy topology on K. Let A be a fuzzy K -algebra with induced fuzzy topology A . Then A is called a fuzzy topological K -algebra if for each a K the self mapping ar : (A, A ) (A, A ) dened by ra (g ) = g is relatively fuzzy continuous. Theorem 4.4. Let f : K1 K2 be a homomorphism of K -algebras. Let and be the fuzzy topologies on K1 and K2 respectively such that = f 1 ( ). If B is a fuzzy topological ideal of K2 with membership function B . Then f 1 (B ) is a fuzzy topological ideal of K1 with membership function f 1 (B) . Proof. It is easy to show that f 1 (B) (e) f 1 (B) (x), for all x K1 . For any x, y K1 , we have f 1 (B) (x) = = B (f (x)) min{B (f (x y )), B (f (y (y x)))} min{B (f (x) f (y )), B (f (y ) f (y x)))} min{f 1 (B) (x y ), f 1(B) (y (y x))}. a g K, (1)

Hence f 1 (B ) is a fuzzy ideal of K1 . Let F1 be an open fuzzy set in f 1 (B) on f 1 (B ). Since f is a fuzzy continuous mapping of (K1 , ) into (K2 , ), f is a relatively fuzzy continuous mapping of ). If there exists open fuzzy set F2 B such that (f 1 (B ), f 1 (B) ) into (B, B f 1 (F2 ) = F1 Then
1 a) ra (F1 ) (g ) = F1 (ra (g ))) = F1 (g = f 1 (F2 ) (g a) = F2 (f (g = F2 (f (g ) f (a))

(2)

a))

1 ra (F1 ) (g ) = F2 (f (g )

f (a))

(3)

Since B is a fuzzy topological in K2 , the mapping rb : (B, B ) (B, B ) dened by rb (y ) = y b b K2 (4)

On fuzzy topological K -algebras

1123

is relatively fuzzy continues.

1 ra (F1 ) (g ) = F2 (f (g )

f (a)) = F2 (rf (a) (f (g )))

1 1 = r1 f (a)(F2 ) (f (g )) = f (rf (a) (F2 ))(g ) 1 1 (F1 ) = f 1 (rf This implies that ra (a) (F2 )) and 1 1 1 1 so ra (F1 ) f (B ) = f (rf (a) (F2 )) f 1 (B ) is open in the induced fuzzy topology on f 1 (B ).

Theorem 4.5. Let f : K1 K2 be an isomorphism of K -algebras. Let and be the fuzzy topologies on K1 and K2 respectively such that f ( ) = . If A is a fuzzy topological ideal of K1 with membership function A . Then f (A) is a fuzzy topological ideal of K2 with membership function f (A) . Proof. It is easy to show that f (A) (e) f (A) (x), for all x K2 . Given x, y K2 , let x0 f 1 (x), y0 f 1 (y ) such that A (x0 )= suptf 1 (x) (t), A (y0 )= suptf 1 (y) (t). Then, we have f (A) (x) =
tf 1 (x)

sup A (t) x0 ))} A (t)} x))}.

A (x0 ) min{A (x0 = min{ sup

tf 1 (x y )

y0 ), A (y0 (y0 A (t), sup y ), f (A)(y

tf 1 (y (y x))

= min{f (A) (x

(y

Hence f (A) is a fuzzy ideal of K2 . Now we show that the mapping rb : (f (A), f (A) ) (f (A), f (A) ) dened by rb (y ) = y b (5) is relatively fuzzy continuous for each b K2 . Let UA be a fuzzy set in A . Then there exists a fuzzy set U in such that UA = U A. Since f is one-one, it follows that f (UA ) = f (U A) = f (U ) f (A) which is a fuzzy set in f (A) . This shows that f is relatively fuzzy open. Let Vf (A) be a fuzzy set in f (A) . The onto mapping of f implies that for each b K2 , there exists a K1 such that b = f (a) Thus f 1 (r1 (Vf (A) ) (x) = f 1 (r1
b f (a)

(6)

(Vf (A) ) (x)

= r 1

f (a)

(Vf (A) ) (f (x))

= Vf (A) (rf (a) (f (x))) = Vf (A) (f (x) f (a)) = f 1 (Vf (A) ) (x a) = f 1 (Vf (A) ) (ra (x))
1 1 (f = ra (Vf (A) )(x).

1124

M. Akram and K. H. Dar

1 This implies that f 1 (ra (Vf (A) )) = r 1 (f 1 (Vf (A) )). The mapping ra : (A, A ) (A, A ) dened by ra (x) = x a is relatively fuzzy continuous and f is rela tively fuzzy continuous map (A, A ) (f (A), f (A) ). 1 1 1 1 Hence f (rb (Vf (A) )) A = ra (f (Vf (A) )) A is a fuzzy set in A .

References
[1] L. A. Zadeh, Fuzzy sets, Information Control, 8(1965) 338-353. [2] C. Chang, Fuzzy topological space, J. Math. Anal. Appl., 24(1968), 182190. [3] A. Rosenfeld, Fuzzy groups, J. Math. Anal. Appl. 35 (1971), 512517. [4] R. Luwen, Fuzzy topological spaces and fuzzy compactness, J. Math. and Anal. Appl., 56(1976), 621-633. [5] D. H. Foster, Fuzzy topological groups, J. Math. Anal. Appl. 67 (1979), 549564. [6] D. M. Ali and A. K. Srivastava , On fuzzy connectedness , Fuzzy Sets and Systems, 28(1988), 203-208. [7] A. K. Chaudhuri and P. Das, Fuzzy connected sets in fuzzy topological spaces, Fuzzy Sets and Systems, 49(1992), 223-229. [8] Y. B. Jun, Fuzzy topological BCK-algebras, Math. Japon. 38(6)(1993), 1059-1063. [9] K. H. Dar and M. Akram, On a K -algebra built on a group, SEA Bull. Math. 29(1) (2005), 41-49. [10] K. H. Dar and M. Akram, Characterization of a K(G)-algebra by self maps, SEA Bull. Math. 28(4) (2004), 601-610. Received: October 3, 2005