Conversely, assume that (G; , e) is a p-semisimple BCI-algebra. If we define
x + y := x (e y), then (G; +) is an abelian group and x − y = x y (see
[4]). Hence we can prove: (K3) x x = x − x = e; (K4) x e = x − e = x;
(K2) x (x y) = x − (x − y) = y, since (G; +) is an abelian group; (K1)
(xy)(xz)=(x−y)−(x−z) = z −y = z y; (K5) ex = e−x = −x,
which is an inverse of x ∈ X. Thus (G; ·, , e) is a K-algebra.
By using Theorem 3.8 and Meng’s observation [4] we conclude:
Proposition