2017 12th International Conference on Intelligent Systems and Knowledge Engineering (ISKE)
A similarity-based fuzzy soft reasoning method
Lu Wang
Southwest Jiaotong University
College of Mathematics
Chengdu,611756, China
965151752@qq.com
Binbin Xue
Southwest Jiaotong University
College of Mathematics
Chengdu,611756, China
xuebinbin@yeah.net
Keyun Qin
Southwest Jiaotong University
College of Mathematics
Chengdu,611756, China
keyunqin@263.net
Abstract—The soft set theory proposed by Molodtsov is
regarded as an effective mathematical tool for dealing with
uncertainties. This paper discusses the approximate reasoning
based on the similarity measure for fuzzy soft set. The ap-
proximate reasoning methods based on the similarity measure
for fuzzy soft modus ponens (FSMP ) are investigated. The
corresponding computational formulas for these methods are
presented. Furthermore, the reductive properties and the
monotonic properties of these methods are analyzed.
Keywords-Similarity measure; Fuzzy soft set; Implication
function; t-Norm;
I. INTRODUCTION
In order to solve the problems in economics, engineering,
environmental science, and social science, the traditional
methods in mathematics are not always successful, because
these problems exist in all kinds of uncertainties. Soft set
theory was initially proposed by Molodtsov in 1999 [1]. It
can be regarded as a new mathematical tool for dealing with
uncertainties and applied in various fields such as decision
making, data analysis, forecasting and texture classification
[2, 3, 4]. The researches of soft set theory are progressing
rapidly. Maji et al. [5] and Ali et al. [6] investigated the
algebraic structures and operations for soft sets. Qin et al.
[7] introduced the definition of soft equality. Meanwhile,
lattice structures and soft quotient algebras of soft sets are
established. Maji et al. [8] studied the hybrid structures
involving soft sets and fuzzy sets. The thought of fuzzy
soft set is proposed, which is the fuzzy generalization of
the classic soft set, and some basic properties are discussed.
Later, many researchers have worked on the concept. Various
kinds of extended fuzzy soft sets such as generalized fuzzy
soft sets [9], intuitionistic fuzzy soft sets [10, 11], interval-
valued fuzzy soft sets [12], vague soft sets [13], interval-
valued intuitionistic fuzzy soft sets [14] and soft interval set
[15, 16] were presented.
The measurement of uncertainty is an important topic for
the theories dealing with uncertainty. As a method to de-
scribe the similarity between two fuzzy sets, the basic theory
of similarity measure has been successful applied to various
fields. Li [17] introduced three approaches to construct
similarity measures of interval-valued fuzzy sets. In [18],
Liu introduced the axioms of similarity measure and entropy
for fuzzy soft set and a category of similarity measures
and entropies is presented based on fuzzzy equivalence.
There is a closely link between the similarity degree and
fuzzy reasoning. In fuzzy reasoning, Zadeh introduced the
method of Composition Rule of Inference(CRI). Although
it has been successful in various systems, the researchers
have pointed out that its mechanism has some drawbacks.
This led to the introduction of similar approximations rea-
soning method. The similarity-based approximate reasoning
method is a simple method which can draw conclusions
that coincide with the truth in some cases. Turksen et al.
[19] presented the concept of similarity measure can be
successfully applied in approximate reasoning. Then, they
proposed the approximate reasoning method for similarity
measure called approximate analogical reasoning schema
(AARS). In [20, 21], Chen proposed two approximate rea-
soning methods based on similarity measure for medical
diagnosis. Chun [22] proposed a method of bidirectional
approximation reasoning based on similarity measure which
is proposed to express the character of decision maker.
Wang et al. [23] introduced the fuzzy similarity reasoning
method and applied it to the framework of fuzzy logic. In
the above work, the approximate reasoning method based
on similarity measure is not necessary to establish fuzzy
relation. The result of the rule, the degree of membership of
each element of the fuzzy set, is corrected by the similarity
between the fact and the rule. In [24], Raha et al. presented
a set of axioms which calculate a reasonable similarity
measure between two imprecise concepts expressed as fuzzy
sets. Meanwhile, Raha et al. calculated a similarity index
between the fact and the antecedent of a rule and used
them in the inference mechanism. In [25], In the fuzzy
environment of interval-value, Feng and Liu proposed a
new approximation reasoning method based on similarity
measure . The research shows that the main mechanism of
the Feng’s method is similar to the Rara’s method, and only
uses different schemes to calculate the modified conditional
relation.
After Zadeh introduced the notion of fuzzy sets, various
methods of fuzzy reasoning have been presented and they
have been used as formal mathematical tools for reasoning
under vagueness. We note that, up to now, there is very little
978-1-5386-1829-5/17/$31.00
c
2017 IEEE