Elsevier

Information Sciences

Volume 129, Issues 1–4, November 2000, Pages 211-220
Information Sciences

Invariant characters of information systems under some homomorphisms

Communicated by Zdzislaw Pawlak
https://doi.org/10.1016/S0020-0255(00)00017-7Get rights and content

Abstract

An information system is one of the very important mathematical models in the field of artificial intelligence. The notion of an information system homomorphism as a kind of tool to study the relation between two information systems was introduced in J.W. Grzymala-Busse (Algebraic properties of knowledge representation systems, in: Proceedings of the ACM SIGART International Symposium on Methodologies for Intelligent Systems, Knoxville, 1986, pp. 432–440). In this paper, some invariant characters of an information system under a homomorphism are discussed.

Introduction

The concept of an information system (also called a knowledge representation system), with manipulations, based on rough set theory was introduced by Pawlak in [1]. This concept and its approach arose from studying problems of expert system, inductive reasoning, learning theory, cluster analysis, and so on. Many topics on an information system have been widely investigated by many scholars (see [2], [3], [4], [5], [6]).

The notion of a homomorphism of information systems as a kind of tool to study the relationship between two information systems was introduced by Grzymala-Busse in [3]. In [5], the authors depicted the conditions which make an information system S=(U,Q,V,f) to be selective (i.e., for arbitrary x∈U[x]Q={x}) in terms of endomorphisms of S. The concept of a homomorphism on information systems is very useful for aggregating sets of objects, attributes, and descriptors of the original system.

In this paper, our main aim is to discuss some invariant characters of an information system under a homomorphism on it. At the same time, we shall give a sufficient condition guaranteeing the existence of an endomorphism of an information system.

In the following section, we recall some terms and definitions to be used in the paper. In Section 3, we establish some lemmas for the sake of the verification of our main results in Section 4. In Section 5, we employ an example to illustrate our main results. Section 6 concludes the paper.

Section snippets

Information systems

By an information system S, we mean an ordered quadrupleS=(U,Q,V,f),where U is a set called the universe of S – elements of  U are called objects, Q is a set of attributes, V(=⋃aQVa) is a set of values of attributes – Va will be called the domain of a, f:U×QV is a description function, such that f(x,a)∈Va for every aQ and xU.

We say that objects x,yU are indiscernible with respect to aQ iff f(x,a)=f(y,a), and we shall write xãy. It is clear that ã is an equivalence relation on U. Objects x,yU

Lemmas

Lemma 3.1

Let S=(U,Q,V,f) and S=(U,Q,V,f) be two information systems and (ho,hA,hD) a homomorphism of S into S. Then ho([x]P)⊆[ho(x)]hA(P) for arbitrary subset PQ and arbitrary xU. In practice, if ho is a surjection and hD is one-to-one, then the inverse inclusion is available, namely, ho([x]P)=[ho(x)]hA(P).

Proof

For arbitrary yho([x]P), there exists an element y∈[x]P such that ho(y)=y. By the definition of a homomorphism, for any aP, it follows that f(ho(y),hA(a))=hD(f(y,a))=hD(f(x,a))=f(ho(x),hA

Theorems

Theorem 4.1

Let S=(U,Q,V,f) be an information system. Let PQ, QP be superfluous in Q, and for each aQP, there exists aP such that ãã, namely, a dependent on a. Denote the information system (U,P,∪aPVa,f) by S, where f is the restriction of f to set U×P. Then, there exists an endomorphism h of S into S.

Proof

If P=Q, it is obvious that the theorem is true. So we can assume that PQ.

We define a mapping h: SS as follows:h=(ho,hA,hD),whereho=1U(theidentitymappingonU),hA(a)=aifa∈P,a(ãã)ifa∈Q−P.

An example

In this section, we employ an example to illustrate our main results.

Let S=(U,Q,V,f) and S=(U,Q,V,f) be two information systems described in Table 3, Table 4, respectively.

Let ho be a mapping from U onto U defined as following:

x1,x5x3,x6x2x8x4x7x9
y1y2y3y4y5y6y7

Let hA be a mapping from Q to Q defined as following:hA(t)=t,wheret∈Q={a,b,c,d}.

Let hD be a mapping from V to V that consists of ha,hb,hcandhd. They are defined by following:ht(vti)=vtit=a,b,c,d;i=1,2,3.

It is very easy to verify

Conclusions

We know that the theory of rough sets deals with the approximation of an arbitrary subset of a universe by two definable or observable subsets called lower and upper approximations (see [1,7]). However, lower and upper approximations are not primitive notions. They are constructed from other concepts, such as binary relations on a universe, partitions and coverings of a universe, and approximation space. For an information system, we can see it as a mixture with some approximation spaces on the

Acknowledgements

This work has been supported by the National Youth Science Foundation of China under Grant 69805004 and the Youth Science Foundation of Shanxi under Grant 981002. The authors thank the referee very much for his or her comments.

References (10)

  • Z. Pawlak

    Rough classification

    Int. J. Man-Mach. Stud.

    (1984)
  • Z. Pawlak

    Rough sets

    Int. J. Inform. Comput. Sci.

    (1982)
  • J.W. Grzymala-Busse, Algebraic properties of knowledge representation systems, in: Proceedings of the ACM SIGART...
  • Z. Pawlak

    On superfluous attributes in knowledge representation system

    Bull. Pol. Acad. Sci. Tech. Sci.

    (1984)
  • J.W. Grzymala-Busse et al.

    On rough sets, and information system homomorphisms

    Bull. Pol. Acad. Sci. Tech. Sci.

    (1988)
There are more references available in the full text version of this article.

Cited by (60)

  • Information structures in a fuzzy set-valued information system based on granular computing

    2021, International Journal of Approximate Reasoning
    Citation Excerpt :

    [48] Li et al. [19] put forward the concept of homomorphism between ISs as below. [19]

View all citing articles on Scopus
View full text