1.

Introduction

While modeling the problems of real world, several concepts have vagueness and ambiguity in the
data obtained and analyzed for many purposes . There is a rapid increase in this vagueness, as well
as the amount and sophistication of such information. All the mathematical notions are specific and
exact in classical mathematics. Classical mathematical tools, therefore, are not always effective in
addressing problems with uncertainties. Over the years, in order to deal with these uncertainties,
the scientists and researchers have been trying to find some suitable instruments. The most popular
are fuzzy set theory [1], soft set theory [7] and rough set theory [9,10]. These theories have
magically reduced the distance between conventional mathematical designs and ambiguous real-
world results.

The fuzzy set theory [1] tests the level of uncertainty of the object data. A fuzzy sets η in a nonempty
universe U of objects is identified by a membership function η :T →[0,1]. This function assigns to
each object t ∈ T , a value of membership η(t ) specifying the degree to which t pertains the
property of η . But, the value 1-η(t ) does not specify the degree to which t has the counter property
of η . For example, A fuzzy set can better define the brightness by associating a degree (between 0
and 1), say 0.7, to a cosmetic object. But, in that object the lightness may not be equal to 0.3. This
could be less than 0.3 or even zero. This implies that  it is not always possible for a fuzzy set η to
function to quantify the degree to which an object has the counter property of η . To tackle this
bipolarity of information, Zhang [3] introduced the bipolar fuzzy sets in 1994. In a nonempty
universe T, a bipolar fuzzy set η is defined by a pair of membership functions as η={¿. The positive
membership function η+¿ :T →[0,1 ]¿ measures the satisfaction degree of the objects with regards to the
property of η, while the function of negative membership η−¿:T →[−1,0] ¿ tests the satisfaction degree
of those objects with regard to some implicit counter-property of η . These sets demonstrate the
degree to which the cosmetic items having bright shade, as well as, the degree to which the
cosmetic item having light shade. For example, let a bipolar fuzzy set η=¿), defining the bright shade
of cosmetic object, equate the value (0.7,-0.1) with a specific cosmetic object `t'. It implies that the
degree of bright shades of the cosmetic object `t' is 0.7 and the degree of light shades in `t' is -0.1.
Indeed, the bipolar fuzzy sets have ability to manage such bipolarity of the information, and thus
have strong impacts on many areas.

The new concept of soft sets, a new mathematical method for dealing with inaccuracies and
uncertainties, was initiated by Molodtsov [7]. A soft set define the objects that have a specific
property or attribute. The soft sets, however, are also not sufficient method to deal with this
bipolarity of information. For example, a piece of jewelry that is not costly may not inherently
inexpensive. In order to solve this problem, the bipolar soft sets were introduced by Shabir and Naz
[8] introduced the bipolar soft sets, which beautifully addressed the bipolarity of information.

Pawlak [9,10] presented the rough set theory with a different approach to resolve the ambiguity in
the information. To investigate how close the objects are to the information attached to them,
Pawlak used the upper and lower approximations of a set of products. These approximations
describe the definite and uncertain sections in the membership values assigned to the objects.

Though the Pawlak model uses an equivalence relations to indiscernibility of the variable.
Nonetheless, equivalence relations are not always possible in the real world because of few
Circumstances. Many researchers therefore put forward generalizations that make use of a Non-
equivalence relation or a covering. We were also witnesses of the birth for several generalized
models for rough set model, including binary relation based rough sets, neighborhood rough sets,
covering based rough sets, rough fuzzy sets and fuzzy rough sets, probabilistic rough sets, etc.

2. Literature Survey

In 1990, Dubois and Prade [18] hybridized rough sets(briefly, RSs) with fuzzy sets(briefly, FSs) to
introduce the notions of rough fuzzy sets(briefly, RFSs) and fuzzy rough sets(briefly, FRSs). Feng et al.
[15] combined soft sets with fuzzy sets and rough sets in a very fruitful way. Zhang [3] introduced the
bipolar fuzzy sets(briefly, BFSs), as an extension of FSs. Lee [14] compared the BFSs with
intuitionistic FSs and interval-valued FSs. Maji et al. [11] presented an application of soft sets(briefly,
SSs) in a decision-making problem. After that, Shabir and Naz applied bipolar soft sets(briefly, BSSs)
in [8] and fuzzy bipolar soft sets(briefly, FBSSs) in [13] to solve decision-making problems. Ma et al.
[12] made a survey of decision-making methodologies based on certain hybrid soft set models.

Covering based rough sets(briefly, CRSs) are among the extended models that are studied most in RS
theory. In 1983, Zakowski [2] proposed a straight forward covering based generalization of the
Pawlak approximation operator. A more detailed description of the advances on CRS appear in
several recent papers such as Yao [6] and D'eer et al. [16, 17]. By combining FS with CRSs, we
generalize CRSs to covering based fuzzy rough sets(briefly, CFRSs). D'eer et al. [16, 17] discussed
fuzzy neighborhoods based on CFRSs.

3. Motivation

In [4, 18] the decisions are made in the favour of an object by using the membership values assigned
to an object. Since these membership values are the upshots of estimation, the definiteness or
uncertainty of these values is not confirmed. This surely affects the reliability and quality of the
decision made. Secondly, the algorithms presented in [4,18] provide the best decision but do not
point out the worst decision. While, in real world problems, There are circumstances in which, even
though the best decision is understood, one might not be able to make the best decision. So, there is
a need to identify the worst decision as well, in order to avoid the worst decision. Multicriteria
decision making can be viewed as a key tool for evaluating plant diseases information to diagnose on
the basis of symptoms and signs, but sometimes symptoms are not obvious and show bipolarity in
behavior. In this situation the BFSs can be of the best choice.

4. Objectives

Our main objectives of research in this area are to introduce the notions of covering based rough
bipolar fuzzy soft sets(briefly, CRBFSSs) rough bipolar fuzzy sets and rough fuzzy bipolar soft
sets(briefly, RFBSSs), by investigating roughness in different bipolar structures, subject to the
available information. We also intend to design algorithms which also provide the worst decision,
along with the best decision and which are capable to accommodate the opinions of more than one
decision makers.
 5. Methodology

We shall investigate the structural properties of the covering based rough bipolar fuzzy sets(briefly,
CRBFSs), rough fuzzy bipolar soft sets(briefly, CRFBSSs) and rough bipolar fuzzy soft sets (briefly,
CRBFSSs)by defining their lower and upper rough approximations, firstly by using a monotone
covering and then by using a bipolar fuzzy covering of a universe set. After that, we shall discuss
some applications of covering based rough bipolar fuzzy sets, rough fuzzy bipolar soft sets and rough
bipolar fuzzy soft sets in some decision-making problems.