1. bookAHEAD OF PRINT
Journal Details
License
Format
Journal
eISSN
2444-8656
First Published
01 Jan 2016
Publication timeframe
2 times per year
Languages
English
Open Access

Research on rule extraction method based on concept lattice of intuitionistic fuzzy language

Published Online: 20 May 2022
Volume & Issue: AHEAD OF PRINT
Page range: -
Received: 04 Mar 2022
Accepted: 10 Apr 2022
Journal Details
License
Format
Journal
eISSN
2444-8656
First Published
01 Jan 2016
Publication timeframe
2 times per year
Languages
English
Introduction

The rapid progress of artificial intelligence has accelerated the development of society. Nowadays, people punch in and out of work using fingerprint identification, which avoids the need to wait in a queue for a long time; such an arrangement not only saves each worker's time but also ensures that the company does not need to maintain or rely on any manual record of attendance, and that absentees, latecomers and early leavers can be known at a glance. Scanning code payment and scanning code collection are achieved everywhere by a mobile phone [1], which saves people a lot of unnecessary trouble and makes their lives tend to be simplified and convenient. At the same time, ‘talking’ robots can sing, tell jokes and chat with people, which relieves people's mental stress [2]. Therefore, the development of intelligence has promoted the growth of human happiness index and the continuous progress of society.

In the field of artificial intelligence research, to make the research results more beneficial to people's lives, it is necessary that the computer be able to not only handle the calculation of general numerical information but also manage the data processing of intelligent information (that is, non-numerical information) close to the human daily thinking mode [3]. In the natural environment, most things that people come into contact with are uncertain, and in the process of reasoning and decision, more emphasis is placed on the use of natural languages with strong fuzziness, such as ‘a little low’, ‘absolutely true’ and ‘very high’ [4]. Therefore, the development of intelligent information processing is particularly important. It can penetrate into any part of people's daily life, and moreover, computers can be used to imitate the thinking mode of the human brain, which accelerates the development of artificial intelligence [5].

Concept lattice, as an effective mathematical tool, can effectively handle knowledge processing and data mining [6, 7], which are widely used in the research of reasoning decision, machine learning, recommendation system, intelligent evaluation, human factor engineering and other fields [8]. Concept lattice is a formal concept model composed of intension and extension that was first proposed by R Wille in 1982 [9]. Nowadays, it is being widely used in the fields of knowledge representation, data mining and machine learning [10, 11].

Extracting implicit and potential rules from information systems is the key step to achieve knowledge acquisition. In the traditional expert system or fuzzy reasoning system, rules are often given by experts according to their experience, which may lead to some problems [12] owing to the fact that the rules acquired based on classical rough set theory are often clear rules with steep truncation, and compared with fuzzy rules, their generalisation ability and practicability are weak [13]. Intuitionistic fuzzy sets (IFS) constitute an important extension of fuzzy sets; their mathematical description is more in line with the nature of fuzzy objects in the objective world, and IFS have thus become a new research hotspot in recent years [14]. Based on retaining the membership function of fuzzy set, IFS adds a new attribute parameter – non-membership function [15], which can further describe the ‘fuzzy concept of either this or the other’; that is, membership, non-membership and hesitation of IFS can represent the three states of support, opposition and neutrality, respectively [16].

Context and attribute reduction of intuitionistic fuzzy language

Concept lattice is a very effective tool in the field of decision analysis. In real life, people are more accustomed to expressing information with linguistic values rather than numerical values [17], and sometimes consider problems with both pros and cons aspects. In order to better deal with linguistic values expressed in terms of both positive and negative aspects, concept lattice theory is used to deal with IFS and fuzzy linguistic values. Moreover, concept lattice of intuitionistic fuzzy language is established [18], and a method of attribute reduction is proposed in the context of intuitionistic fuzzy language.

Intuitionistic fuzzy language concept lattice
Definition 1

Define (U,A,S) as an formal context of intuitionistic fuzzy language, in which U = {x1, x2, ⋯ , xn} is an object set, A = {a1, ⋯ , am} is attribute set, S is binary relation between U and A, and SU × A; regard S(x,a)=sμ(x,a),sν(x,a) {S^{'}}(x,a) = {s_\mu}(x,a),{s_\nu}(x,a) as the intuitionistic fuzzy language value of object X under attribute A, where sμ:U×A{s0,s1,,sg},sv:U×A{s0,s1,,sg}and0μ+vg. {s_\mu}:U \times A \to \left\{{{s_0},{s_1}, \cdots,{s_g}} \right\},\quad {s_v}:U \times A \to \left\{{{s_0},{s_1}, \cdots,{s_g}} \right\}{\kern 1pt} {\rm{and}}{\kern 1pt} 0 \le \mu + v \le g.

Further, sμ(x,a) and sν(x,a) represent the linguistic value membership degree and linguistic value non-membership degree between object X and attribute A, respectively [19]. As ∀x,yU, ∀a,bA, there are: S(x,a)S(y,b)(sμ(x,a)sμ(y,b))(sv(x,a)sv(y,b))S(x,a)c=sv(x,a),sμ(x,a) \matrix{{{S^{'}}(x,a) \ge {S^{'}}(y,b) \Leftrightarrow \left({{s_\mu}(x,a) \ge {s_\mu}(y,b)} \right) \cap \left({{s_v}(x,a) \le {s_v}(y,b)} \right)} \cr {{S^{'}}{{(x,a)}^c} = {s_v}(x,a),{s_\mu}(x,a)} \cr}

Definition 2

Define (U,A,S) as an intuitionistic fuzzy language formal context, where ∀XU, SA S_A^{'} represents a set of object subset X under attribute set A, BSA B \subseteq S_A^{'} represents the attribute intuitionistic fuzzy language values, and at the same time, a pair of derived operators is defined as follows: X<={xXsμ(x,a),vxXsv(x,a)aA,xX} {X^ <} = \cap \left\{{\left\langle {\mathop \wedge \limits_{x \in X} {s_\mu}(x,a),\mathop v\limits_{x \in X} {s_v}(x,a)} \right\rangle a \in A,x \in X} \right\} B={xS(x,a)B(x,a),xU,aA} {B^ \prec} = \cap \left\{{x{S^{'}}(x,a) \ge B(x,a),x \in U,\forall a \in A} \right\} in which B(x,a) represents the intuitionistic fuzzy language value of object X under attribute A.

Definition 3

Define (U,A,S) as an intuitionistic fuzzy language formal context; if a tuple (X,B) satisfies X = B and X = B, (X,B) is regarded as the concept of intuitionistic fuzzy language form, in which X is the extension of intuitionistic fuzzy language form concept and B is the intension of intuitionistic fuzzy language form concept [20].

Definition 4

Define (U,A,S) as an intuitionistic fuzzy language form context, make IFLL (U,A,S) represent (U,A,S) on the whole intuitionistic fuzzy language concept [21] and IFLL (U,A,S) is called intuitionistic fuzzy language concept lattice, for ∀(X1, B1), (X2, B2) ∈ IFLL (U,A,S), defining its partial order relation as (X1,B1)(X2,B2)X1X2(B1B2) \left({{X_1},{B_1}} \right) \le \left({{X_2},{B_2}} \right) \Leftrightarrow {X_1} \subseteq {X_2}\left({\Leftrightarrow {B_1} \supseteq {B_2}} \right) in which (X1, B2) is called the child concept of (X2, B2) and (X2, B2) is called the parent concept of (X1, B2).

Define the lower and upper bounds of IFLL (U,A,S) as [22] (X1,B1)(X2,B2)=((X1X2),B1B2) \left({{X_1},{B_1}} \right) \cap \left({{X_2},{B_2}} \right) = \left({{{\left({{X_1} \cup {X_2}} \right)}^{\prec \prec}},{B_1} \cap {B_2}} \right) (X1,B1)(X2,B2)=(X1X2,(B1B2)) \left({{X_1},{B_1}} \right) \cup \left({{X_2},{B_2}} \right) = \left({{X_1} \cap {X_2},{{\left({{B_1} \cup {B_2}} \right)}^{\prec \prec}}} \right) where IFLL (U,A,S) is a complete lattice.

Example 1

Given a formal context of intuitionistic fuzzy language (U,A,S) U = (x1, x2, x3), A = (a,b,c) and make the language set S = {s0 = verylow, s1 = low, s2 = medium, s3 = high, s4 = veryhigh}. Among them, S (x1, a) = s0, s4 means the possibility that the object x1 has the attribute A is very low, and the possibility that it does not have the attribute A is very high (Table 1).

Formal context of Intuitionistic fuzzy linguistic (U,A,S)

UAbc

x1s0, s4s2, s2s1, s2
x2s3, s0s0, s2s3, s1
x3s1, s1s3, s1s2, s1

Table 1 shows the formal context of intuitionistic fuzzy language (U,A,S) The generated concept is:

#1 (U, {s0, s4s0, s2s1, s2})

#2 ({x1x3}, {s0, s4s2, s2s1, s2})

#3 ({x2x3}, {s1, s1s0, s2s2, s1})

#4 ({x2}, {s3, s0s0, s2s3, s1})

#5 ({x3}, {s1, s1s3, s1s2, s1})

#6 (∅, {s3, s0s3, s1s3, s1})

The intuitionistic fuzzy language concept lattice is built as shown in Figure 1:

Fig. 1

Intuitionistic fuzzy linguistic concept lattice IFLL (U,A,S)

According to the above definition, the following properties can be obtained:

Property 1, applicable in the context of intuitionistic fuzzy language form (U,A,S): Next, for any XU and BSA B \subseteq S_A^{'} , there are:

XX≺≺, BB≺≺

X = X≺≺≺, B = B≺≺≺

XBBX

(X≺≺, X) and (B, B≺≺) are all concepts.

Property 2, applicable in the context of intuitionistic fuzzy language form (U,A,S): For any X1, X2U and B1,B2SA {B_1},{B_2} \subseteq S_A^{'} , there are:

X1X2X2X1 {X_1} \subseteq {X_2} \Rightarrow X_2^ \prec \subseteq X_1^ \prec , B1B2B2B1 {B_1} \subseteq {B_2} \Rightarrow B_2^ \prec \subseteq B_1^ \prec ;

(X1X2)=X1X2 {\left({{X_1} \cup {X_2}} \right)^ \prec} = X_1^ \prec \cap X_2^ \prec , (B1B2)=B1B2 {\left({{B_1} \cup {B_2}} \right)^ \prec} = B_1^ \prec \cap B_2^ \prec ;

(X1X2)X1X2 {\left({{X_1} \cap {X_2}} \right)^ \prec} \supseteq X_1^ \prec \cup X_2^ \prec , (B1B2)B1B2 {\left({{B_1} \cap {B_2}} \right)^ \prec} \supseteq B_1^ \prec \cup B_2^ \prec

Theorem 1

If (X1, B1) and (X2, B2) are two concepts of the formal context of intuitionistic fuzzy language (U,A,S), then X1X2,(B1B2)≺≺) and ((X1X2)≺≺, B1B2) are all concepts.

Prove

It is evidently proved by the above nature and definition.

Attribute reduction in the formal context of intuitionistic fuzzy language

In the era of big data, the amount of information is huge. Under this context, the construction of concept lattice is very complicated, and it is thus necessary to reduce the attributes of concept lattice [23]. In this section, the related definitions and properties of attribute reduction in the context of intuitionistic fuzzy language form are obtained [24].

Definition 5

Define IFLL(U,A1,S1) IFLL\left({U,{A_1},S_1^{'}} \right) and IFLL(U,A2,S2) IFLL\left({U,{A_2},S_2^{'}} \right) as two conceptual lattices of intuitionistic fuzzy language; if (X,B)IFLL(U,A2,S2) \forall \left({{X^{'}},{B^{'}}} \right) \in IFLL\left({U,{A_2},S_2^{'}} \right) , (X,B)IFLL(U,A1,S1) \exists (X,B) \in IFLL\left({U,{A_1},S_1^{'}} \right) so that X = X, which is called IFLL(U,A1,S1) IFLL\left({U,{A_1},S_1^{'}} \right) , IFLL(U,A2,S2) IFLL\left({U,{A_2},S_2^{'}} \right) as: IFLL(U,A1,S1)IFLL(U,A2,S2) IFLL\left({U,{A_1},S_1^{'}} \right) \le IFLL\left({U,{A_2},S_2^{'}} \right)

If IFLL(U,A1,S1)IFLL(U,A2,S2) IFLL\left({U,{A_1},S_1^{'}} \right) \le IFLL\left({U,{A_2},S_2^{'}} \right) and IFLL(U,A1,S1)IFLL(U,A2,S2) IFLL\left({U,{A_1},S_1^{'}} \right) \ge IFLL\left({U,{A_2},S_2^{'}} \right) are established, it is called IFLL(U,A1,S1) IFLL\left({U,{A_1},S_1^{'}} \right) , and IFLL(U,A2,S2) IFLL\left({U,{A_2},S_2^{'}} \right) is isomorphic and regarded as IFLL(U,A1,S1)IFLL(U,A2,S2) IFLL\left({U,{A_1},S_1^{'}} \right) \cong IFLL\left({U,{A_2},S_2^{'}} \right) .

Under the formal context of intuitionistic fuzzy language (U,A,S), ∀DA and SD=S(U×D) S_D^{'} = {S^{'}} \cap (U \times D) ; then, (U,D,SD) \left({U,D,S_D^{'}} \right) is also a formal context of intuitionistic fuzzy language [25]. IFLL(U,D,SD) IFLL\left({U,D,S_D^{'}} \right) refers to the formal context of expressing intuitionistic fuzzy language (U,D,SD) \left({U,D,S_D^{'}} \right) Under the concept lattice, for operators X(XU), (X,B)IFLL(U,D,SD) \forall (X,B) \in IFLL\left({U,D,S_D^{'}} \right) , satisfy: if aD, XD (a) = X(a).

Theorem 2

Assuming (U,A,S) is the formal context of intuitionistic fuzzy language [26], (X,B)IFLL(U,D,SD) \forall (X,B) \in IFLL(U,D,S_D^{'}) , DA yes BBD.

Proof

It can be easily proved by property 1 and definition 5.

Theorem 3

Assuming (U,A,S) is the formal context of intuitionistic fuzzy language, forDA, D ≠ ∅ there is always IFLL(U,A,S)IFLL(U,D,SD) IFLL\left({U,A,{S^{'}}} \right) \le IFLL\left({U,D,S_D^{'}} \right) .

Prove

(X,B)IFLL(U,D,SD) \forall (X,B) \in IFLL\left({U,D,S_D^{'}} \right) , XD = B and BD = X, and as can be seen from property 1, (X≺≺, X) ∈ IFLL (U,A,S), XX<. According to theorem 2, B = XDX ⇒ X≺≺BBD = X. Therefore, IFLL(U,A,S) ≤ IFLL(U,D,Sb).

Definition 6

(U,A,S) is the formal context of intuitionistic fuzzy language; if ∃DA, make IFLL(U,D,SD)IFLL(U,A,S) IFLL\left({U,D,S_D^{'}} \right) \cong IFLL\left({U,A,{S^{'}}} \right) , and then d is a coordination set. If D is a coordination set and IFLL(U,Dd,SDd) IFLL\left({U,D - d,S_{D - d}^{'}} \right) , is different from IFLL (U,A,S), and then D is called (U,A,S). About Jane. All the intersection of (U,A,S)reduction is called the core of (U,A,S).

The following theorems can be obtained through definition 6:

Theorem 4

Assuming (U,A,S) is formal context of intuitionistic fuzzy language, DA, D ≠ ∅ and then D is the coordination set. IFLL(U,D,SD)IFLL(U,A,S) \Leftrightarrow IFLL\left({U,D,S_D^{'}} \right) \le IFLL\left({U,A,{S^{'}}} \right) .

Definition 7

Assumes that all reductions of the formal context of intuitionistic fuzzy language (U,A,S) are {DiDi isareduction, iτ} (τ is an index set) [27]. Attribute set a can be divided into the following four parts:

Absolutely necessary attribute (core attribute) b: biτDi b \in \mathop \cap \limits_{i \in \tau} {D_i}

Relative essential attribute C: ciτDiiτDi c \in \mathop \cup \limits_{i \in \tau} {D_i} - \mathop \cup \limits_{i \in \tau} {D_i}

Absolutely unnecessary attribute d: dAiτDi d \in A - \mathop \cup \limits_{i \in \tau} {D_i}

Unnecessary attribute E: eAiτDi e \in A - \mathop \cap \limits_{i \in \tau} {D_i}

Theorem 5

Under the formal context of any intuitionistic fuzzy language (U,A,S), there must be at least one reduction.

Prove

If for any aA, there are IFLL(U,A{a},SA{a}) IFLL\left({U,A - \{a\},S_{A - \{a\}}^{'}} \right) different from IFLL (U,A,S), then A itself is a reduction.

If it exists aA, make IFLL(U,A{a},SA{a})=IFLL(U,A,S) IFLL\left({U,A - \{a\},S_{A - \{a\}}^{'}} \right) = IFLL\left({U,A,{S^{'}}} \right) ; if for any a1A − {a}, there are IFLL(U,A{a,a1},SA{a,a1}) IFLL\left({U,A - \left\{{a,{a_1}} \right\},S_{A - \left\{{a,{a_1}} \right\}}^{'}} \right) different from IFLL (U,A,S), then A−{a1} is about Jane; if IFLL(U,A{a,a1},SA{a,a1})IFLL(U,A,S) IFLL\left({U,A - \left\{{a,{a_1}} \right\},S_{A - \left\{{a,{a_1}} \right\}}^{'}} \right) \cong IFLL\left({U,A,{S^{'}}} \right) , A − {a,a1} needs to be discussed. Since A is a finite set, at least one reduction will be found.

Generally, the reduction intuition blurs the formal context of (U,A,S) is not unique.

Example 2

For the formal context of intuitionistic fuzzy language shown in Table 2 (U,A,S), U = {x1, x2, x3, x4}, A = {a,b,c,d,e} and the concept lattice of intuitionistic fuzzy language and its reduction are obtained.

Intuitionistic fuzzy linguistic for mal context (U,A,S)

Uabcde

x1s4, s0s3, s1s1, s2s3, s0s3, s0
x2s3, s0s3, s1s3, s0s1, s2s1, s3
x3s0, s3s1, s2s0, s4s3, s1s1, s2
x4s3, s1s3, s0s2, s0s1, s2s1, s3

The formal context of intuitionistic fuzzy language shown in Table 2 has nine concepts:

#1 ({U}, {s0, s3s1, s2s0, s4s1, s2s1, s3})

#2 ({x1x2x4}, {s3, s1s3, s1s1, s2s1, s2s1, s3})

#3 ({x1x2}, {s3, s0s3, s1s1, s2s1, s2s1, s3})

#4 ({x1x3}, {s0, s3s1, s2s0, s4s3, s1s1, s2})

#5 ({x2x4}, {s3, s1s3, s1s2, s0s1, s2s1, s3})

#6 ({x1}, {s4, s0s3, s1s1, s2s4, s0s3, s0})

#7 ({x2}, {s3, s0s3, s1s3, s0s1, s2s1, s3})

#8 ({x4}, {s3, s1s3, s0s2, s0s1, s2s1, s3})

#9 (∅, {s3, s1s3, s1s1, s2s1, s2s1, s3})

Its intuitionistic fuzzy language concept lattice IFLL (U,A,S) is shown in Figure 2.

Fig. 2

Intuitionistic fuzzy linguistic concept lattice IFLL (U,A,S)

Define D1 = {a,b,c,d}, in the context of intuitionistic fuzzy language form (U,D1,SD1) \left({U,{D_1},S_{{D_1}}^{'}} \right) ; the generated concept is

#1 (U, {s0, s3s1, s2s0, s4s1, s2})

#2 ({x1x2x4}, {s3, s1s3, s1s1, s2s1, s2})

#3 ({x1x2}, {s3, s0s3, s1s1, s2s1, s2})

#4 ({x1x3}, {s0, s3s1, s2s0, s4s3, s1})

#5 ({x2x4}, {s3, s1s3, s1s2, s0s1, s2})

#6 ({x1}, {s4, s0s3, s1s1, s2s4, s0})

#7 ({x2}, {s3, s0s3, s1s3, s0s1, s2})

#8 ({x4}, {s3, s1s3, s0s2, s0s1, s2})

#9 (∅, {s3, s1s3, s1s1, s2s1, s2})

Its intuitionistic fuzzy language concept lattice IFLL(U,D1,SD1) IFLL\left({U,{D_1},S_{{D_1}}^{'}} \right) is shown in Figure 3.

Fig. 3

Intuitionistic fuzzy linguistic concept lattice IFLL(U,D1,SD1) IFLL\left({U,{D_1},S_{{D_1}}^{'}} \right)

As can be seen from Figures 2 and 3, IFLL(U,D1,SD1)IFLL(U,A,S) IFLL\left({U,{D_1},S_{{D_1}}^{'}} \right) \cong IFLL\left({U,A,{S^{'}}} \right) but does not exist. d1D1 to make IFLL(U,D1{d1},SD1(d1))IFLL(U,A,S) IFLL\left({U,{D_1} - \left\{{{d_1}} \right\},S_{{D_1} - \left({{d_1}} \right)}^{'}} \right) \cong IFLL\left({U,A,{S^{'}}} \right) , and thus D1 = {a,b,c,d} is a reduction of (U,A,S). Similarly, D2 = {a,b,c,e} is a reduction of (U,A,S).

To sum up, intuitionistic fuzzy language formal context of (U,A,S) has two reductions: D1 = {a,b,c,d} and D2 = {a,b,c,e}. That is, the core attributes are a, b and c, the relative necessary attributes are d and e, and there are no unnecessary attributes [28].

Corollary 1

Assuming that (U,A,S) is the formal context of intuitionistic fuzzy language, then the core attribute is reduction.The reduction is the only one.

Corollary 2

Assuming that (U,A,S) is the formal context of intuitionistic fuzzy language, then aA is an unnecessary attribute.A − {a} is a coordination set [29].

Corollary 3

Assuming that (U,A,S) is the formal context of intuitionistic fuzzy language, then aA is a core attribute.A − {a} is not a coordination set.

The reduction D on the concept lattice of intuitionistic fuzzy language needs to satisfy the following two conditions:

D is a coordination set;

dD, D − {d} is not a coordination set.

Therefore, for attribute reduction, it is necessary to determine whether the attribute subset is a coordination set, and the judgment theorem of the coordination set is given as follows:

Theorem 6

It is assumed (U,A,S) that formal context of intuitionistic fuzzy language, DA, D ≠ ∅. So, D is the coordination setB≺≺DD = B, BSA \forall B \subseteq S_A^{'} .

Prove

Necessity: Assuming that D is a coordinated set, according to theorem 4, there is IFLL(U,D,SD)IFLL(U,A,S) IFLL\left({U,D,S_D^{'}} \right) \le IFLL(U,A,{S^{'}}) , and for BSA \forall B \subseteq S_A^{'} , (B, B≺≺) ∈ IFLL (U,A,S). According to definition 5, BSA \exists {B^{'}} \subseteq S_A^{'} satisfies (B,B)IFLL(U,D,SD) \left({{B^ \prec},{B^{'}}} \right) \in IFLL(U,D,S_D^{'}) . Therefore, B = B≺≺D and □ = □′≺□; therefore B≺≺DD = B.

Sufficiency: BSA \forall B \subseteq S_A^{'} , B≺≺DD = B. Asuming that (X,B) ∈ IFLL(U,A,S), that is, there is X = B, B = X is established. Take B = B≺≺D, B′≺D = B≺≺DD = B = X. Therefore, (X,B)IFLL(U,D,SD) (X,{B^{'}}) \in IFLL(U,D,S_D^{'}) , that is, for ∀(X,B) ∈ IFLL(U,A,S), BSA \exists B' \subseteq S_A^{'} satisfies (X,B)IFLL(U,D,SD) (X,{B^{'}}) \in IFLL(U,D,S_D^{'}) , and thus D is the coordination set.

Corollary 4

Assuming that (U,A,S) is the formal context of intuitionistic fuzzy language, DC, D ≠ ∅. So D is the coordination set BSA \forall B \subseteq S_A^{'} , BSA \exists {B^{'}} \subseteq S_A^{'} satisfy B′≺D = B.

Theorem 7

Assuming that (U,A,S) is the formal context of intuitionistic fuzzy language, DA, D ≠ ∅, E = AD. So, D is the coordination set BSE \forall B \subseteq S_E^{'} , BSD \exists {B^{'}} \subseteq S_D^{'} satisfy B′≺D = BE.

Proof

Necessity: D is the coordination set of (U,A,S); then, for B1SA \forall {B_1} \subseteq S_A^{'} , B1=B1D B_1^ \prec = B_1^{{\prec _D}} . With regard to BSE \forall B \subseteq S_E^{'} , set B2 = BB1, in which B1SD {B_1} \subseteq S_D^{'} , B1=BB B_1^ \prec = {B^{{\prec _B}}} . According to corollary 4, BSD \exists B \subseteq S_D^{'} satisfies B′≺D = BE.

Sufficiency: For BSE \forall B \subseteq S_E^{'} , BSD \exists B \subseteq S_D^{'} satisfies B<D = BE. Suppose (X,B2) ∈ IFLL (U,A,S), then X = B2 and X=B2 X = B_2^ \prec . Assume that B2 = BB1, in which BSE B \subseteq S_E^{'} , B1SD {B_1} \subseteq S_D^{'} ; then B2 = (BB1) = BB1 = BDB1E. Therefore BSD \exists {B^{'}} \subseteq S_D^{'} satisfies BD=B1E {B^{'{\prec _D}}} = B_1^{\prec E} , and therefore BDB1E = B1DB′≺D = (B1 ∪ B)D = X. ((B1B)<D,(B1B)<D<D)IFLL(U,D,SD) \left({{{\left({{B_1} \cup {B^{'}}} \right)}^{{< _D}}},{{\left({{B_1} \cup {B^{'}}} \right)}^{{< _D}{< _D}}}} \right) \in IFLL\left({U,D,S_D^{'}} \right) , and thus IFLL(U,D,SD)IFLL(U,A,S) IFLL\left({U,D,S_D^{'}} \right) \le IFLL\left({U,A,{S^{'}}} \right) .

To sum up, D is the coordination set of (U,A,S).

Attribute reduction of intuitionistic fuzzy language concept lattice based on discernibility matrix

Discernibility matrix is a common tool for attribute reduction. The identification matrix and identification function are used to reduce the concept lattice of intuitionistic fuzzy language [30].

Definition 8

Assume that (U,A,S) is the formal context of intuitionistic fuzzy language, (Xi,Bi) and (Xj,Bj), and (Xi,Bi),(Xj,Bj) ∈ IFLL (U,A,S). Then, D˜((Xi,Bi),(Xj,Bj))={{aAμX1(a)>μXj(a)orvXi(a)<vXj(a)},XiXj,Xi,XjU,else \tilde D\left({\left({{X_i},{B_i}} \right),\left({{X_j},{B_j}} \right)} \right) = \left\{{\matrix{{\left\{{a \in A{\mu _{{X_1}}}(a) > {\mu _{{X_j}}}(a)or{v_{{X_i}}}(a) < {v_{{X_j}}}(a)} \right\},{X_i} \not\subset {X_j},{X_i},{X_j} \ne U,\emptyset} \hfill \cr {\emptyset \quad {\rm{else}}} \hfill \cr}} \right.

D˜R=(D˜((Xi,Bi),(Xj,Bj))(Xj,Bj)),(Xi,Bi),(Xj,Bj)IFLL(U,A,S)) {\tilde D^R} = \left. {\left({\tilde D\left({\left({{X_i},{B_i}} \right),\left({{X_j},{B_j}} \right)} \right)\left({{X_j},{B_j}} \right)} \right),\left({{X_i},{B_i}} \right),\left({{X_j},{B_j}} \right) \in IFLL\left({U,A,{S^{'}}} \right)} \right) is regarded as an identifiable matrix on the formal context of intuitionistic fuzzy language.

Theorem 8

Assume that (U,A,S) is the formal context of intuitionistic fuzzy language, DA, D ≠ ∅, ∀ (Xi,Bi), (Xj,Bj) ∈ IFLL (U,A,S), (Xi,Bi) ≠ (Xj,Bj); then the following two propositions are equivalent:

D is the coordination set.

If D˜((Xi,Bi),(Xj,Bj)) \tilde D\left({\left({{X_i},{B_i}} \right),\left({{X_j},{B_j}} \right)} \right) \ne \emptyset , then DD˜((Xi,Bi),(Xj,Bj)) D\bigcap\nolimits_ \tilde D\left({\left({{X_i},{B_i}} \right),\left({{X_j},{B_j}} \right)} \right) \ne \emptyset , D˜((Xi,Bi),(Xj,Bj))D˜R \tilde D\left({\left({{X_i},{B_i}} \right),\left({{X_j},{B_j}} \right)} \right) \in {\tilde D^R}

Proof

(1) (2): Assuming that (2) does not hold true, D˜((Xi,Bi),(Xj,Bj))D˜R \exists \tilde D\left({\left({{X_i},{B_i}} \right),\left({{X_j},{B_j}} \right)} \right) \ne \emptyset \in {\tilde D^R} , Xi ⊄ Xj satisfy DD˜((Xi,Bi),(Xj,Bj))= D\bigcap\nolimits_ \tilde D\left({\left({{X_i},{B_i}} \right),\left({{X_j},{B_j}} \right)} \right) = \emptyset . That is, ∀aD satisfies μ Xi(a) > μ Xj(a).

Definition 9

Assume that (U,A,S) is the intuitionistic fuzzy linguistic form context; D*R is the identification matrix on (U,A,S); define f(D˜R)=D˜((Xi,Bi),(Xj,Bj))D˜R{akD˜((Xi,Bi),(Xj,Bj))ak}D˜((Xi,Bi),(Xj,Bj) \matrix{{f(\tilde DR) = \mathop {\tilde D(({X_i},{B_i}),({X_j},{B_j})) \in {{\tilde D}^R}}\limits^ \wedge \left\{{\mathop {{a_k} \in \tilde D(({X_i},{B_i}),({X_j},{B_j}))}\limits^ \vee {a_k}} \right\}} \cr {\tilde D(({X_i},{B_i}),({X_j},{B_j}) \ne \emptyset} \cr}

Then, f(D˜R) f\left({{{\tilde D}^R}} \right) is the identification function of (U,A,S).

Theorem 9

Assume that (U,A,S) is the formal context of intuitionistic fuzzy language, and the minimal disjunctive normal form of the discernibility function is defined as Λ(D˜R)=pk=1[qkarr=1] \Lambda \left({{{\tilde D}^R}} \right) = \matrix{p \cr \vee \cr {k = 1} \cr} \left[ {\matrix{{{q_k}} \cr {\wedge {a_r}} \cr {r = 1} \cr}} \right]

Make Bk = {arr = 1,2, ⋯ , qk}; then, {Bkk = 1,2, ⋯ , p} are all reductions of intuitionistic fuzzy language form context (U,A,S).

It can be seen from the above that to find the attribute reduction of the conceptual lattice of intuitionistic fuzzy language, the following conditions need to be obtained:

The smallest coordination set D of DD˜((Xi,Bi),(Xj,Bj))= D\bigcap\nolimits_ \tilde D\left({\left({{X_i},{B_i}} \right),\left({{X_j},{B_j}} \right)} \right) = \emptyset , and D˜((Xi,Bi),(Xj,Bj))= \tilde D\left({\left({{X_i},{B_i}} \right),\left({{X_j},{B_j}} \right)} \right) = \emptyset .

Corollary 5

D is (U,A,S) coordination set onX0DXα<.

Table 3 gives the intuitionistic fuzzy language form context of (U,A,S) identifiable matrix.

Discernible matrix of Intuitionistic fuzzy linguistic formal context (U,A,S)

/□124121324124

1adeadeade
2cacabc
4bcbbcabc
12aabca
13dedededede
24ccabc
124abc

Note. Sinceis not considered in the process of attribute reduction, theis omitted in the identification matrix. Λ(D˜*)=(ade)(bc)(ac)(abc)(de)abc=abc(de)=(abcd)(abce) \matrix{{\Lambda \left({{{\tilde D}^*}} \right)} \hfill & {= (a \vee d \vee e) \wedge (b \vee c) \wedge (a \vee c) \wedge (a \vee b \vee c) \wedge (d \vee e) \wedge a \wedge b \wedge c = a \wedge b \wedge c \wedge (d \vee e)} \hfill \cr {} \hfill & {= (a \wedge b \wedge c \wedge d) \vee (a \wedge b \wedge c \wedge e)} \hfill \cr}

The formal context of intuitionistic fuzzy language shown in Table 3 has two reductions, namely: and. Among them, a, b and c, are core essential attributes; d and e are relatively necessary attributes. These observations are consistent with the conclusion of example 2.

Rule extraction method based on intuitionistic fuzzy language concept lattice

Rule extraction is an important branch in the field of concept lattice. In the era of big data, knowledge acquired from complex data is largely given in the form of rules [31]. Based on the formal context of intuitionistic fuzzy language, this paper puts forward the formal context of intuitionistic fuzzy language decision, studies the acquisition method of rules, discusses the similarity between intuitionistic fuzzy language sets and constructs a rule extraction method based on concept lattice of intuitionistic fuzzy language.

Rule extraction under the decision formal context of intuitionistic fuzzy language
Definition 10

Assume that (U,A,S, T,R) is a decision formal context of intuitionistic fuzzy language, in which (U,A,S) and (U,T,R) are the context of intuitionistic fuzzy language, AT = ∅, S is Intuitionistic fuzzy language binary relation between U ×A, R is intuitionistic fuzzy language binary relation between U × T, A is called conditional attribute set and T is called decision attribute set.

Under the decision formal context of intuitionistic fuzzy language (U,A,S, T,R), RT R_T^{'} A set of intuitionistic fuzzy language values representing the object set U under the decision attribute set T [32].

Definition 11

Assuming that (U,A,S, T,R) is an intuitionistic fuzzy language decision formal context, if IFLL (U,A,S) ≤ IFLL (U,T,R), the context is regarded to be coordinated.

IFLL (U,A,S) is a conditional concept lattice of intuitionistic fuzzy language, whereas IFLL (U,T,R) is a decision concept lattice for intuitionistic fuzzy language.

Definition 12

Assuming that (U,A,S, T,R) is a decision formal context of intuitionistic fuzzy language, IFLL (U,A,S) ≤ IFLL (U,T,R); if (X,B) ∈ IFLL (U,A,S), (Y,C) ∈ IFLL (U,T,R), satisfy X = Y (XU, ∅), it is called B→ C as the rule derived under the context of decision, which is recorded as if B, then C, and all the rules are recorded in Φ in (a).

Similarity of language values in the context of intuitionistic fuzzy language decision formal
Definition 13

Assuming that (U,A,S) is a formal context of intuitionistic fuzzy language, for ∀XU B,GSA B,G \subseteq S_A^{'} , (X,B) ∈ IFLL (U,A,S) then the distance between two intuitionistic fuzzy language sets B = {sα1, sβ1 sα2, sβ2sαm, sβm} and G = {sγ1, sλ1 sγ2, sλ2sγm, sλm} is: d(B,G)=1ml=1m|αlγl|+|βlλl|2g d(B,G) = {1 \over m}\mathop {\sum\limits_{l = 1}^m}{{\left| {{\alpha _l} - {\gamma _l}} \right| + \left| {{\beta _l} - {\lambda _l}} \right|} \over {2g}}

Among them, l = 1, ⋯ , m.

Theorem 10

Assuming that B = {sα1, sβ1 sα2, sβ2sαm, sβm}, G = {sγ1, sλ1 sγ2, sλ2sγm, sλm} for any two intuitionistic fuzzy language sets, the distance d(B,G) of language sets in the context of intuitionistic fuzzy language forms satisfies the following properties:

0 ≤ d(B,G) ≤ 1;

d(B,G) = 0 if and only if B = G B=G;

d(B,G) = d(G,B);

Assume Q = {〈sκ1, sρ1〉, 〈sκ2, ρ2〉 ⋯ 〈sκm, sρm〉} as an intuitionistic fuzzy language set, BGQ, then d(B,G) ≤ d(B,Q), d(G,Q) ≤ d(B,Q).

Proof

αllll ∈ [0, g]

|αlγl| ∈ [0, g],|βlλl| ∈ [0, g],|αlγl| + |βlλl| ∈ [0,2g],

then 0|αlγl|+|βlλl|2g101nl=1m|αlγl|+|βlλl|2g1 0 \le {{\left| {{\alpha _l} - {\gamma _l}} \right| + \left| {{\beta _l} - {\lambda _l}} \right|} \over {2g}} \le 1 \Rightarrow 0 \le {1 \over n}\sum\limits_{l = 1}^m {{\left| {{\alpha _l} - {\gamma _l}} \right| + \left| {{\beta _l} - {\lambda _l}} \right|} \over {2g}} \le 1

d(B,G)=01nl=1m|αlγl|+|βlλl|2g=0 d(B,G) = 0 \Leftrightarrow {1 \over n}\sum\limits_{l = 1}^m {{\left| {{\alpha _l} - {\gamma _l}} \right| + \left| {{\beta _l} - {\lambda _l}} \right|} \over {2g}} = 0

|αlγl|+|βlλl|2g=0 \Leftrightarrow {{\left| {{\alpha _l} - {\gamma _l}} \right| + \left| {{\beta _l} - {\lambda _l}} \right|} \over {2g}} = 0

⇔ |αlγl| = 0 and |βlλl| = 0

αl = γl and βl = λl

B = G

d(B,G)=1nl=1m|αlγl|+|βlλl|2g=1nl=1m|γlαl|+|λlβl|2g=d(G,B) d(B,G) = {1 \over n}\sum\limits_{l = 1}^m {{\left| {{\alpha _l} - {\gamma _l}} \right| + \left| {{\beta _l} - {\lambda _l}} \right|} \over {2g}} = {1 \over n}\sum\limits_{l = 1}^m {{\left| {{\gamma _l} - {\alpha _l}} \right| + \left| {{\lambda _l} - {\beta _l}} \right|} \over {2g}} = d(G,B)

If BGQ, yes αlγlκ1, βlλlρl, |αlγl| ≤ |αlκl|, |βlλl| ≤ |βlρl|, then d(B,G)=1nl=1m|αlγl|+|βlλl|2g1nl=1m|αlκl|+|βlρl|2g=d(B,Q) d(B,G) = {1 \over n}\sum\limits_{l = 1}^m {{\left| {{\alpha _l} - {\gamma _l}} \right| + \left| {{\beta _l} - {\lambda _l}} \right|} \over {2g}} \le {1 \over n}\sum\limits_{l = 1}^m {{\left| {{\alpha _l} - {\kappa _l}} \right| + \left| {{\beta _l} - {\rho _l}} \right|} \over {2g}} = d(B,Q)

Similarly, d(GQ)≤ d(B,Q) can be proved.

To sum up, theorem 10 is realised.

Definition 14

Assuming that (U,A,S) is a formal context of intuitionistic fuzzy language, for ∀XU, B,GSA B,G \subseteq S_A^{'} , (X,B) ∈ IFLL (U,A,S), then the similarity between two intuitionistic fuzzy languages is B = {〈sα1, sβ1〉 〈 sα2, sβ2〉 ⋯ 〈 sαm, sβm〉} and G = {〈sγ1, sλ1〉 〈sγ2, sλ2〉 ⋯ 〈 sγm, sλm〉}: simII(B,G)=1d(B,G)=11ml=1m|αlγl|+|βlλl|2g {\rm{si}}{{\rm{m}}_{{\rm{II}}}}(B,G) = 1 - d(B,G) = 1 - {1 \over m}\sum\limits_{l = 1}^m {{\left| {{\alpha _l} - {\gamma _l}} \right| + \left| {{\beta _l} - {\lambda _l}} \right|} \over {2g}}

Theorem 11

Assuming that B = {〈sα1, sβ1〉 〈sα2, sβ2〉 ⋯ 〈sαm, sβm〉}, G = {〈sγ1, sλ1〉 〈sγ2, sλ2〉 ⋯ 〈sγm, sλm〉}: are any two intuitionistic fuzzy language sets, then the similarity of language sets in the context of intuitionistic fuzzy language forms. simII(B,G) satisfying the following properties:

0 ≤ simII(B,G) ≤ 1;

simII(B,G) = 1 if and only if B=G;

simII(B,G) = simII(G,B);

Assume Q = {〈sκ1, sρ1〉, 〈sκ2, ρ2〉 ⋯ 〈sκm, sρm〉} as any intuitionistic fuzzy language set, BGQ, then simII(B,G) ≥ simII(B,Q), simII(G,Q) ≥ simII(B,Q).

Definition 15

Assuming that (U,A,S) is a formal context of intuitionistic fuzzy language, for ∀XU, B,GSA B,G \subseteq S_A^{'} , (X,B) ∈ IFLL (U,A,S), wII = (w1, w2, …,wm)T is the weight vector then the weighted distance between two intuitionistic fuzzy language sets B = {〈sα1, sβ1〉 〈sα2, sβ2〉 ⋯ 〈sαm, sβm〉} and G = {〈sγ1, sλ1〉 〈sγ2, sλ2〉 ⋯ 〈sγm, sλm〉} is: dw(B,G)=l=1mwl(|αlγl|+|βlλl|)2g {d_w}(B,G) = \sum\limits_{l = 1}^m {{{w_l}\left({\left| {{\alpha _l} - {\gamma _l}} \right| + \left| {{\beta _l} - {\lambda _l}} \right|} \right)} \over {2g}} where wl ∈ [0,1] and l=1mwl=1(l=1,,m) \sum\limits_{l = 1}^m {w_l} = 1(l = 1, \cdots,m) .

Theorem 12

Assuming B = {〈sα1, sβ1〉 〈sα2, sβ2〉 ⋯ 〈sαm, sβm〉}, G = {〈sγ1, sλ1〉 〈sγ2, sλ2〉 ⋯ 〈sγm, sλm〉} as any two intuitionistic fuzzy language sets, wII = (w1, w2, …,wm) is the weight vector, where wl ∈ [0,1] and l=1mwl=1(l=1,,m) \sum\limits_{l = 1}^m {w_l} = 1(l = 1, \cdots,m) , and the weighted distance of language set in the context of language form is intuitively fuzzy. dw(B,G) satisfies the following properties:

0 ≤ dw(B,G) ≤ 1;

dw(B,G)=0 if and only if B=G;

dw(B,G)=dw(GB);

Assume Q = {〈sκ1, sρ1〉, 〈sκ2, ρ2〉 ⋯ 〈sκm, sρm〉} as any intuitionistic fuzzy language set, BGQ, then: dw(B,G) ≤ dw(B,Q), dw(G,Q) ≤ dw(B,Q).

Definition 16

Assuming that (U,A,S) is a formal context of intuitionistic fuzzy language, for ∀XU, B,GSA B,G \subseteq S_A^{'} , (X,B) ∈ IFLL (U,A,S), wII = (w1, w2, …,wm)T is the weight vectorthen, the weighted similarity between two intuitionistic fuzzy language sets B = {〈sα1, sβ1〉 〈sα2, sβ2〉 ⋯ 〈sαm, sβm〉} and G = {〈sγ1, sλ1〉 〈sγ2, sλ2〉 ⋯ 〈sγm, sλm〉} is: simwI(B,G)=1dw(B,G)=1l=1nωl(|αlγl|+|βlλl|)2g {\rm{si}}{{\rm{m}}_{{w_{\rm{I}}}}}(B,G) = 1 - {d_w}(B,G) = 1 - \sum\limits_{l = 1}^n {{{\omega _l}\left({\left| {{\alpha _l} - {\gamma _l}} \right| + \left| {{\beta _l} - {\lambda _l}} \right|} \right)} \over {2g}} where wl ∈ [0,1] and l=1mwl=1(l=1,,m) \sum\limits_{l = 1}^m {w_l} = 1(l = 1, \cdots,m) .

Theorem 13

Assuming that B = {〈sα1, sβ1〉 〈sα2, sβ2〉 ⋯ 〈sαm, sβm〉}, G = {〈sγ1, sλ1〉 〈sγ2, sλ2〉 ⋯ 〈sγm, sλm〉}: are any two intuitionistic fuzzy language sets, wII = (w1, w2, …,wm) is the weight vector, where wl ∈ [0,1] and l=1mwl=1(l=1,,m) \sum\limits_{l = 1}^m {w_l} = 1(l = 1, \cdots,m) . Then the weighted similarity of language sets in the context of intuitionistic fuzzy language forms simwII(B,G) satisfying the following properties:

0 ≤ simwI(B,G) ≤ 1;

simwII(B,G) = 1 if and only if B=G;

simwI(B,G) = simwI(G,B);

Assume Q = {〈sκ1, sρ1〉 〈sκ2, ρ2〉 ⋯ 〈sκm, sρm〉} as any intuitionistic fuzzy language set, BGQ, then:

simWI(B,G)simWI(B,Q),simWII(G,Q)simWI(B,Q) {\rm{si}}{{\rm{m}}_{{W_{\rm{I}}}}}(B,G) \ge {\rm{si}}{{\rm{m}}_{{W_{\rm{I}}}}}(B,Q),{\rm{si}}{{\rm{m}}_{{W_{{\rm{II}}}}}}(G,Q) \ge {\rm{si}}{{\rm{m}}_{{W_{\rm{I}}}}}(B,Q)

Note. When each element in the weight vector is evenly distributed, that is w1=w2==wm=1m {w_1} = {w_2} = \ldots = {w_m} = {1 \over m} , the weighted similarity degenerates into similarity, which means that similarity is a special case of weighted similarity.

A rule extraction method based on concept lattice of intuitionistic fuzzy language

Under the context of intuitionistic fuzzy language decision formal, language rules are extracted. First, the conditional concept lattice and decision concept lattice are constructed; then, according to the detailed relationship between concept lattices, the rule set under the context of intuitionistic fuzzy language decision formal is obtained. Finally, the given intuitionistic fuzzy language set is compared with each rule antecedent in the rule set, and according to the formula of weighted similarity, the rules of a given intuitionistic fuzzy language set are judged. The specific steps are as follows:

Step 1: Generate all concepts (X,B) on the formal context (U,A,S) of intuitionistic fuzzy language to construct conditional concept lattice. IFLL (U,A,S);

Step 2: Generate all concepts (Y,C) on the formal context (U,T,R) of intuitionistic fuzzy language to construct decision concept lattice. IFLL (U,T,R);

Step 3: Judge the relationship between the two concept lattices IFLL (U,A,S) and IFLL (U,T,R). According to X = Y, all rules under the context of intuitionistic fuzzy language decision form are obtained, and the satisfied rules are added to the rule set. Φ(A);

Step 4: Construct a given intuitionistic fuzzy language set: G={sγ1,sλ1sγ2,sλ2sγm,sλm} G = \left\{{\left\langle {{s_{{\gamma _1}}},{s_{{\lambda _1}}}} \right\rangle \left\langle {{s_{{\gamma _2}}},{s_{{\lambda _2}}}} \right\rangle \ldots \left\langle {{s_{{\gamma _m}}},{s_{{\lambda _m}}}} \right\rangle} \right\}

Combine intuitionistic fuzzy language set G with rule set Φ(A). Compare the antecedent B of each rule in this set; if there is B=G and B→ C, the algorithm ends and G→ C is output; if there is no B=G, proceed to step 5;

Step 5: According to the weighted similarity formula: simwI(B,G)=1l=1mωl(|αlγl|+|βlλl|)2g {\rm{si}}{{\rm{m}}_{{w_{\rm{I}}}}}(B,G) = 1 - \sum\limits_{l = 1}^m {{{\omega _l}\left({\left| {{\alpha _l} - {\gamma _l}} \right| + \left| {{\beta _l} - {\lambda _l}} \right|} \right)} \over {2g}}

Calculate the weighted similarity between intuitionistic fuzzy language set G and antecedent B in each rule, so as to obtain the maximum weighted similarity;

Step 6: Compare the maximum weighted similarity with the similarity threshold; if it is greater than or equal to the similarity threshold, then the rule B→ C corresponding to the maximum weighted similarity is the rule to be searched, then G→ C, the algorithm ends, and G→ C is output. If the maximum weighted similarity is less than the similarity threshold, return to step 1 and add the context until the matching rule is found.

Examples of disease diagnosis based on rule extraction method of intuitionistic fuzzy language concept lattice

By the symptoms and diseases of some patients, the possibility of a patient with certain symptoms suffering from these diseases can be judged. Assuming that (U,A,S, T,R) is a formal context of intuitionistic fuzzy language decision, U = {x1, x2, x3, x4} represents patient set, A = {a,b,c,d,e} represents symptom set and T = {t1, t2, t3, t4} represents disease set. The specific definitions of each of the symptoms under set A are as follows: a is fever, b is headache, c is stomach-ache, d is cough and e is chest pain. The specific definition of disease T is as follows: t1 is viral fever, t2 is typhoid fever, t3 is gastropathy and t4 is thoracic lung disease. The language set S = {s0 = extremely few, s1 = few, s2 = medium, s3 = many, S4 = extremely many} indicates the frequency of certain symptoms in patients, and the linguistic set R = { r0 = small, r1 = small, r2 = medium, r3 = large, r4 = extremely large} indicates that patients suffer from certain diseases. The formal context of intuitionistic fuzzy language decision is shown in Table 4.

Intuitionistic fuzzy linguistic formal decision context (U,A,S, T,R)

abcdet1t3t3t4

x1s3, s1s2, s2s0, s4s2, s2s2, s1r3, r0r2, r2r0, r4r2, r1
x2s4, s0s2, s1s4, s0s3, s0s3, s1r4, r0r3, r1r4, r0r3, r1
x3s1, s2s3, s1s2, s1s0, s3s1, s2r1, r2r2, r2r2, r1r1, r2
x4s2, s2s4, s0s3, s1s4, s0s0, s3r3, r1r4, r0r3, r0r0, r3

Step 1: Construct a conditional concept lattice of intuitionistic fuzzy language;

According to Table 4, the concepts under the context of (U,A,S) form are as follows:

#1 (∅,{〈s4, s0〉, 〈s4, s0〉 〈s4, s0〉 〈s4, s0〉 〈s3, s1〉}),

#2 ({x2}, {〈s4, s0〉 〈s2, s1〉 〈s4, s0〉 〈s3, s0〉 〈s3, s1〉}),

#3 ({x3}, {〈s1, s2〉 〈s3, s1〉 〈s2, s1〉 〈s0, s3〉 〈s1, s2〉}),

#4 ({x4}, {〈s2, s2〉 〈s4, s0〉 〈s3, s1〉 〈s4, s0〉 〈s0, s3〉}),

#5 ({x1x2}, {〈s3, s1〉 〈s2, s2〉 〈s0, s4〉 〈s2, s2〉 〈s2, s1〉}),

#6 ({x2x3}, {〈s1, s2〉 〈s2, s1〉 〈s2, s1〉 〈s0, s3〉 〈s1, s2〉}),

#7 ({x2x4}, {〈s2, s2〉 〈s2, s1〉 〈s3, s1〉 〈s3, s0〉 〈s0, s3〉}),

#8 ({x3x4}, {〈s1, s2〉 〈s3, s1〉 〈s2, s1〉 〈s0, s3〉 〈s0, s3〉}),

#9 ({x1x2x3}, {〈s1, s2〉 〈s2, s2〉 〈s0, s4〉 〈s0, s3〉 〈s1, s2〉}),

#10 ({x1x2x4}, {〈s2, s2〉 〈s2, s2〉 〈s0, s4〉 〈s2, s2〉 〈s0, s3〉}),

#11 ({x2x3x4}, {〈s1, s2〉 〈s2, s1〉 〈s2, s1〉 〈s0, s3〉 〈s0, s3〉}),

#12 (U,{〈s1, s2〉 〈s2, s2〉 〈s0, s4〉 〈s0, s3〉 〈s0, s3〉}).

Construct conditional concept lattice of intuitionistic fuzzy language according to concepts IFLL (U,A,S), as indicated in Figure 4.

Fig. 4

Intuitionistic fuzzy linguistic conditional concept lattice IFLL (U,A,S)

Step 2: Construct the concept lattice of intuitionistic fuzzy language decision;

The concepts under the formal context (U,T,R) are as follows:

#1 (∅, {〈r4, r0〉 〈r4, r0〉 〈r4, r0〉 〈r3, r1〉}), #2 ({x2}, {〈r4, r0〉 〈r3, r1〉 〈r4, r0〉 〈r3, r1〉})

#3 ({x4}, {〈r3, r1〉 〈r4, r0〉 〈r3, r0〉 〈r0, r3〉}), #4 ({x1x2}, {〈r3, r0〉 〈r2, r2〉 〈r0, r4〉 〈r2, r1〉}),

#5 ({x2x3}, {〈r1, r2〉 〈r2, r2〉 〈r2, r1〉 〈r1, r2〉}), #6 ({x2x4}, {〈r3, r1〉 〈r3, r1〉 〈r3, r0〉 〈r0, r3〉})

#7 ({x1x2x3}, {〈r1, r2〉 〈r2, r2〉 〈r0, r4〉 〈r1, r2〉}), #8 ({x1x2x4}, {〈r3, r1〉 〈r2, r2〉 〈r0, r4〉 〈r0, r3〉})

#9 ({x2x3x4}, {〈r1, r2〉 〈r2, r2〉 〈r2, r1〉 〈r0, r3〉}), #10 (U, {〈r1, r2〉 〈r2, r2〉 〈r0, r4〉 〈r0, r3〉})

Construct intuitionistic fuzzy language decision concept lattice according to concepts IFLL (U,T,R) as indicated in Figure 5.

Fig. 5

Intuitionistic fuzzy linguistic decision concept lattice IFLL (U,T,R)

Step 3: Obtain the rule set. Φ(A).

As can be seen from Figures 4 and 5, IFLL (U,A,S) ≤ IFLL (U,T,R), and the rule set can be made available as follows: Φ(A)={{s4,s0s2,s1s4,s0s3,s0s3,s1}{r4,r0r3,r1r4,r0r3,r1},{s2,s2s4,s0s3,s1s4,s0s0,s3}{r3,r1r4,r0r3,r0r0,r3},{s3,s1s2,s2s0,s4s2,s2s2,s1}{r3,r0r2,r2r0,r4r2,r1},{s1,s2s2,s1s2,s1s0,s3s1,s2}{r1,r2r2,r2r2,r1r1,r2},{s2,s2s2,s1s3,s1s3,s0s0,s3}{r3,r1r3,r1r3,r0r0,r3},{s1,s2s2,s2s0,s4s0,s3s1,s2}{r1,r2r2,r2r0,r4r1,r2},{s2,s2s2,s2s0,s4s2,s2s0,s3}{r3,r1r2,r2r0,r4r0,r3},{s1,s2s2,s1s2,s1s0,s3s0,s3}{r1,r2r2,r2r2,r1r0,r3}}. \matrix{{\Phi (A) =} \hfill & {\left\{{\left\{{\left\langle {{s_4},{s_0}} \right\rangle \left\langle {{s_2},{s_1}} \right\rangle \left\langle {{s_4},{s_0}} \right\rangle \left\langle {{s_3},{s_0}} \right\rangle \left\langle {{s_3},{s_1}} \right\rangle} \right\}} \right. \to \left\{{\left\langle {{r_4},{r_0}} \right\rangle \left\langle {{r_3},{r_1}} \right\rangle \left\langle {{r_4},{r_0}} \right\rangle \left\langle {{r_3},{r_1}} \right\rangle} \right\},} \hfill \cr {} \hfill & {\,\,\,\left\{{\left\langle {{s_2},{s_2}} \right\rangle \left\langle {{s_4},{s_0}} \right\rangle \left\langle {{s_3},{s_1}} \right\rangle \left\langle {{s_4},{s_0}} \right\rangle \left\langle {{s_0},{s_3}} \right\rangle} \right\} \to \left\{{\left\langle {{r_3},{r_1}} \right\rangle \left\langle {{r_4},{r_0}} \right\rangle \left\langle {{r_3},{r_0}} \right\rangle \left\langle {{r_0},{r_3}} \right\rangle} \right\},} \hfill \cr {} \hfill & {\,\,\,\left\{{\left\langle {{s_3},{s_1}} \right\rangle \left\langle {{s_2},{s_2}} \right\rangle \left\langle {{s_0},{s_4}} \right\rangle \left\langle {{s_2},{s_2}} \right\rangle \left\langle {{s_2},{s_1}} \right\rangle} \right\} \to \left\{{\left\langle {{r_3},{r_0}} \right\rangle \left\langle {{r_2},{r_2}} \right\rangle \left\langle {{r_0},{r_4}} \right\rangle \left\langle {{r_2},{r_1}} \right\rangle} \right\},} \hfill \cr {} \hfill & {\,\,\,\left\{{\left\langle {{s_1},{s_2}} \right\rangle \left\langle {{s_2},{s_1}} \right\rangle \left\langle {{s_2},{s_1}} \right\rangle \left\langle {{s_0},{s_3}} \right\rangle \left\langle {{s_1},{s_2}} \right\rangle} \right\} \to \left\{{\left\langle {{r_1},{r_2}} \right\rangle \left\langle {{r_2},{r_2}} \right\rangle \left\langle {{r_2},{r_1}} \right\rangle \left\langle {{r_1},{r_2}} \right\rangle} \right\},} \hfill \cr {} \hfill & {\,\,\,\left\{{\left\langle {{s_2},{s_2}} \right\rangle \left\langle {{s_2},{s_1}} \right\rangle \left\langle {{s_3},{s_1}} \right\rangle \left\langle {{s_3},{s_0}} \right\rangle \left\langle {{s_0},{s_3}} \right\rangle} \right\} \to \left\{{\left\langle {{r_3},{r_1}} \right\rangle \left\langle {{r_3},{r_1}} \right\rangle \left\langle {{r_3},{r_0}} \right\rangle \left\langle {{r_0},{r_3}} \right\rangle} \right\},} \hfill \cr {} \hfill & {\,\,\,\left\{{\left\langle {{s_1},{s_2}} \right\rangle \left\langle {{s_2},{s_2}} \right\rangle \left\langle {{s_0},{s_4}} \right\rangle \left\langle {{s_0},{s_3}} \right\rangle \left\langle {{s_1},{s_2}} \right\rangle} \right\} \to \left\{{\left\langle {{r_1},{r_2}} \right\rangle \left\langle {{r_2},{r_2}} \right\rangle \left\langle {{r_0},{r_4}} \right\rangle \left\langle {{r_1},{r_2}} \right\rangle} \right\},} \hfill \cr {} \hfill & {\,\,\,\left\{{\left\langle {{s_2},{s_2}} \right\rangle \left\langle {{s_2},{s_2}} \right\rangle \left\langle {{s_0},{s_4}} \right\rangle \left\langle {{s_2},{s_2}} \right\rangle \left\langle {{s_0},{s_3}} \right\rangle} \right\} \to \left\{{\left\langle {{r_3},{r_1}} \right\rangle \left\langle {{r_2},{r_2}} \right\rangle \left\langle {{r_0},{r_4}} \right\rangle \left\langle {{r_0},{r_3}} \right\rangle} \right\},} \hfill \cr {} \hfill & {\,\,\,\left\{{\left\langle {{s_1},{s_2}} \right\rangle \left\langle {{s_2},{s_1}} \right\rangle \left\langle {{s_2},{s_1}} \right\rangle \left\langle {{s_0},{s_3}} \right\rangle \left\langle {{s_0},{s_3}} \right\rangle} \right\} \to \left. {\left\{{\left\langle {{r_1},{r_2}} \right\rangle \left\langle {{r_2},{r_2}} \right\rangle \left\langle {{r_2},{r_1}} \right\rangle \left\langle {{r_0},{r_3}} \right\rangle} \right\}} \right\}.} \hfill \cr}

Step 4: Judge whether there is a rule precursor in the rule set (a) that is the same as the known patient symptoms; by comparing the rule sets Φ(A) with symptoms of an existing patient, it can be found that there is no rule antecedent with it.

Step 5: Calculate the weighted similarity between the intuitive fuzzy language set G of the patient's symptoms and each rule antecedent in the rule set, so as to obtain the maximum weighted similarity;

Taking the possibility of the patient suffering from viral fever as an example, the weight corresponding to viral fever is given as wII1={0.4,0.05,0.2,0.1,0.25} w_{{\rm{II}}}^1 = \{0.4,0.05,0.2,0.1,0.25\} , assuming that: B1={s4,s0s2,s1s4,s0s3,s0s3,s1}B2={s2,s2s4,s0s3,s1s4,s0s0,s3}B3={s3,s1s2,s2s0,s4s2,s2s2,s1}B4={s1,s2s2,s1s2,s1s0,s3s1,s2}B6={s1,s2s2,s2s0,s4s0,s3s1,s2}B7={s2,s2s2,s2s0,s4s2,s2s0,s3}B8={s1,s2s2,s1s2,s1s0,s3s0,s3} \matrix{{{B_1} = \left\{{\left\langle {{s_4},{s_0}} \right\rangle \left\langle {{s_2},{s_1}} \right\rangle \left\langle {{s_4},{s_0}} \right\rangle \left\langle {{s_3},{s_0}} \right\rangle \left\langle {{s_3},{s_1}} \right\rangle} \right\}} \hfill \cr {{B_2} = \left\{{\left\langle {{s_2},{s_2}} \right\rangle \left\langle {{s_4},{s_0}} \right\rangle \left\langle {{s_3},{s_1}} \right\rangle \left\langle {{s_4},{s_0}} \right\rangle \left\langle {{s_0},{s_3}} \right\rangle} \right\}} \hfill \cr {{B_3} = \left\{{\left\langle {{s_3},{s_1}} \right\rangle \left\langle {{s_2},{s_2}} \right\rangle \left\langle {{s_0},{s_4}} \right\rangle \left\langle {{s_2},{s_2}} \right\rangle \left\langle {{s_2},{s_1}} \right\rangle} \right\}} \hfill \cr {{B_4} = \left\{{\left\langle {{s_1},{s_2}} \right\rangle \left\langle {{s_2},{s_1}} \right\rangle \left\langle {{s_2},{s_1}} \right\rangle \left\langle {{s_0},{s_3}} \right\rangle \left\langle {{s_1},{s_2}} \right\rangle} \right\}} \hfill \cr {{B_6} = \left\{{\left\langle {{s_1},{s_2}} \right\rangle \left\langle {{s_2},{s_2}} \right\rangle \left\langle {{s_0},{s_4}} \right\rangle \left\langle {{s_0},{s_3}} \right\rangle \left\langle {{s_1},{s_2}} \right\rangle} \right\}} \hfill \cr {{B_7} = \left\{{\left\langle {{s_2},{s_2}} \right\rangle \left\langle {{s_2},{s_2}} \right\rangle \left\langle {{s_0},{s_4}} \right\rangle \left\langle {{s_2},{s_2}} \right\rangle \left\langle {{s_0},{s_3}} \right\rangle} \right\}} \hfill \cr {{B_8} = \left\{{\left\langle {{s_1},{s_2}} \right\rangle \left\langle {{s_2},{s_1}} \right\rangle \left\langle {{s_2},{s_1}} \right\rangle \left\langle {{s_0},{s_3}} \right\rangle \left\langle {{s_0},{s_3}} \right\rangle} \right\}} \hfill \cr}

Therefore, simwII2(B1,G)=0.219,simwΠ1(B2,G)=0.606,simwI1(B3,G)=0.581,simwII1(B4,G)=0.688,simwII1(B5,G)=0.625,simwII1(B6,G)=0.756simwI1(B7,G)=0.756,simwI1(B8,G)=0.775 \matrix{{{\rm{si}}{{\rm{m}}_{w{\rm{I}}{{\rm{I}}^2}}}\left({{B_1},G} \right) = 0.219,\quad {\rm{si}}{{\rm{m}}_{{w_{{\Pi ^1}}}}}\left({{B_2},G} \right) = 0.606,\quad {\rm{si}}{{\rm{m}}_{w_{\rm{I}}^1}}\left({{B_3},G} \right) = 0.581,} \hfill \cr {{\rm{si}}{{\rm{m}}_{{w_{{\rm{II}}}}^1}}\left({{B_4},G} \right) = 0.688,\quad {\rm{si}}{{\rm{m}}_{{w_{{\rm{II}}}}^1}}\left({{B_5},G} \right) = 0.625,\quad {\rm{si}}{{\rm{m}}_{{w_{{\rm{II}}}}^1}}\left({{B_6},G} \right) = 0.756} \hfill \cr {{\rm{si}}{{\rm{m}}_{w_{\rm{I}}^1}}\left({{B_7},G} \right) = 0.756,\quad {\rm{si}}{{\rm{m}}_{w_{\rm{I}}^1}}\left({{B_8},G} \right) = 0.775} \hfill \cr}

It can be seen from the above that the weighted similarity between G and rule antecedent Bs is the largest.

Step 6: Output the result.

When the similarity threshold is selected as 0.75, the maximum weighted similarity under viral fever is obtained. When simwI1 (BS,G) = 0.775 > 0.75, the following rule is obtained: {s0,s3s2,s0s0,s3s1,s3s0,s4}{r1,r2} \left\{{\left\langle {{s_0},{s_3}} \right\rangle \left\langle {{s_2},{s_0}} \right\rangle \left\langle {{s_0},{s_3}} \right\rangle \left\langle {{s_1},{s_3}} \right\rangle \left\langle {{s_0},{s_4}} \right\rangle} \right\} \to \left\{{\left\langle {{r_1},{r_2}} \right\rangle} \right\}

That is, the possibility of the patient suffering from viral fever is 〈r1, r2〉. Similarly, the given corresponding weight of typhoid fever wII2={0.3,0.4,0.1,0.1,0.1} w_{{\rm{II}}}^2 = \{0.3,0.4,0.1,0.1,0.1\} , weight corresponding to stomach diseases wII3 = {0.1,0.2,0.5,0.1,0.1}, chest and lung diseases corresponding to the right wII4={0.05,0.05,0.1,0.1,0.7} w_{{\rm{II}}}^4 = \{0.05,0.05,0.1,0.1,0.7\} . After calculation, simWII2(B8,G) {\rm{si}}{{\rm{m}}_{W_{{\rm{II}}}^2}}\left({{B_{\rm{8}}},G} \right) , simwII3 (B6, G) and simwII4 (B7, G) are, respectively, the m aximum weighted similarity; so: {s0,s3s2,s0s0,s3s1,s3s0,s4}{r1,r2r2,r2r0,r4r0,r3} \left\{{\left\langle {{s_0},{s_3}} \right\rangle \left\langle {{s_2},{s_0}} \right\rangle \left\langle {{s_0},{s_3}} \right\rangle \left\langle {{s_1},{s_3}} \right\rangle \left\langle {{s_0},{s_4}} \right\rangle} \right\} \to \left\{{\left\langle {{r_1},{r_2}} \right\rangle \left\langle {{r_2},{r_2}} \right\rangle \left\langle {{r_0},{r_4}} \right\rangle \left\langle {{r_0},{r_3}} \right\rangle} \right\}

That is, the possibility of the patient suffering from viral fever is 〈r1, r2〉, the possibility of typhoid fever is 〈r2, r2〉, the possibility of gastropathy is 〈r0, r4〉 and the possibility of suffering from thoracic lung disease is 〈r0, r3〉.

Taking the threshold as 〈r2, r2〉, that is, when the probability of suffering from a certain disease is greater than or equal to 〈r2, r2〉, it should be noticed. Therefore, the patient should pay attention to typhoid fever.

The above examples show that the results are consistent with the analysis in real life, which indicates that the algorithm is effective and practical. In this paper, intelligent diagnosis with linguistic values in the field of disease diagnosis can better express the description of the disease, reduce data loss and thus get smarter diagnosis results.

Conclusion

Based on the concept lattice of intuitionistic fuzzy language, the rule extraction in the context of intuitionistic fuzzy language decision formal is defined in this paper. In order to handle the rules on the concept lattice, the similarity is defined. In addition, by considering the various influences of each attribute on the object, the weight is added to the similarity, and the weighted similarity is proposed. Finally, using an example connected with disease diagnosis, it is demonstrated that the algorithm proposed in this paper is effective and that the proposed decision method has strong operability and is suitable for practical application.

Fig. 1

Intuitionistic fuzzy linguistic concept lattice IFLL (U,A,S′)
Intuitionistic fuzzy linguistic concept lattice IFLL (U,A,S′)

Fig. 2

Intuitionistic fuzzy linguistic concept lattice IFLL (U,A,S′)
Intuitionistic fuzzy linguistic concept lattice IFLL (U,A,S′)

Fig. 3

Intuitionistic fuzzy linguistic concept lattice 



IFLL(U,D1,SD1′)
IFLL\left({U,{D_1},S_{{D_1}}^{'}} \right)
Intuitionistic fuzzy linguistic concept lattice IFLL(U,D1,SD1′) IFLL\left({U,{D_1},S_{{D_1}}^{'}} \right)

Fig. 4

Intuitionistic fuzzy linguistic conditional concept lattice IFLL (U,A,S′)
Intuitionistic fuzzy linguistic conditional concept lattice IFLL (U,A,S′)

Fig. 5

Intuitionistic fuzzy linguistic decision concept lattice IFLL (U,T,R′)
Intuitionistic fuzzy linguistic decision concept lattice IFLL (U,T,R′)

Formal context of Intuitionistic fuzzy linguistic (U,A,S′)

U A b c

x1 s0, s4 s2, s2 s1, s2
x2 s3, s0 s0, s2 s3, s1
x3 s1, s1 s3, s1 s2, s1

Intuitionistic fuzzy linguistic for mal context (U,A,S′)

U a b c d e

x1 s4, s0 s3, s1 s1, s2 s3, s0 s3, s0
x2 s3, s0 s3, s1 s3, s0 s1, s2 s1, s3
x3 s0, s3 s1, s2 s0, s4 s3, s1 s1, s2
x4 s3, s1 s3, s0 s2, s0 s1, s2 s1, s3

Discernible matrix of Intuitionistic fuzzy linguistic formal context (U,A,S′)

/□ 1 2 4 12 13 24 124

1 ade ade ade
2 c ac abc
4 bc b bc abc
12 a abc a
13 de de de de de
24 c c abc
124 abc

Intuitionistic fuzzy linguistic formal decision context (U,A,S′, T,R′)

a b c d e t1 t3 t3 t4

x1 s3, s1 s2, s2 s0, s4 s2, s2 s2, s1 r3, r0 r2, r2 r0, r4 r2, r1
x2 s4, s0 s2, s1 s4, s0 s3, s0 s3, s1 r4, r0 r3, r1 r4, r0 r3, r1
x3 s1, s2 s3, s1 s2, s1 s0, s3 s1, s2 r1, r2 r2, r2 r2, r1 r1, r2
x4 s2, s2 s4, s0 s3, s1 s4, s0 s0, s3 r3, r1 r4, r0 r3, r0 r0, r3

Cui H, Yue G, Zou L, et al. Multiple multidimensional linguistic reasoning algorithm based on property-oriented linguistic concept lattice [J]. International Journal of Approximate Reasoning, 2020, 131. CuiH YueG ZouL Multiple multidimensional linguistic reasoning algorithm based on property-oriented linguistic concept lattice [J] International Journal of Approximate Reasoning 2020 131 10.1016/j.ijar.2020.11.006 Search in Google Scholar

Gou X, Xu Z, Liao H. Hesitant Fuzzy Linguistic Possibility Degree-Based Linear Assignment Method for Multiple Criteria Decision-Making [J]. International Journal of Information Technology & Decision Making (IJITDM), 2019, 18. GouX XuZ LiaoH Hesitant Fuzzy Linguistic Possibility Degree-Based Linear Assignment Method for Multiple Criteria Decision-Making [J] International Journal of Information Technology & Decision Making (IJITDM) 2019 18 10.1142/S0219622017500377 Search in Google Scholar

Zou L, Feng K, Liu X. Linguistic-Valued Intuitionistic Fuzzy Concept Lattice and Its Application. 2018. ZouL FengK LiuX Linguistic-Valued Intuitionistic Fuzzy Concept Lattice and Its Application 2018 Search in Google Scholar

Yan L, Pei Z, Ren F. Constructing and managing multi-granular linguistic values based on linguistic terms and their fuzzy sets [J]. IEEE Access, 2019, PP(99):1–1. YanL PeiZ RenF Constructing and managing multi-granular linguistic values based on linguistic terms and their fuzzy sets [J] IEEE Access 2019 PP(99):1–1. 10.1109/ACCESS.2019.2948847 Search in Google Scholar

Yang X, Xu Z, Zhou W. Green Credit Product Design Based on Fuzzy Concept Lattice [J]. IEEE Access, 2019. YangX XuZ ZhouW Green Credit Product Design Based on Fuzzy Concept Lattice [J] IEEE Access 2019 10.1109/ACCESS.2019.2943486 Search in Google Scholar

Mieszkowicz-Rolka A, Rolka L. The Concept of Fuzzy Linguistic Labels for Studying Decision Systems in Air Transportation [J]. Procedia Computer Science, 2019, 151:923–928. Mieszkowicz-RolkaA RolkaL The Concept of Fuzzy Linguistic Labels for Studying Decision Systems in Air Transportation [J] Procedia Computer Science 2019 151 923 928 10.1016/j.procs.2019.04.128 Search in Google Scholar

Xian S, Guo H, Chai J, et al. Interval probability hesitant fuzzy linguistic analytic hierarchy process and its application in talent selection [J]. Journal of Intelligent and Fuzzy Systems, 2020, 39(1):1–19. XianS GuoH ChaiJ Interval probability hesitant fuzzy linguistic analytic hierarchy process and its application in talent selection [J] Journal of Intelligent and Fuzzy Systems 2020 39 1 1 19 10.3233/JIFS-190427 Search in Google Scholar

Konecny J. Note on m-polar fuzzy graph representation of concept lattice: How to really compute m-polar fuzzy concepts [J]. Engineering Applications of Artificial Intelligence, 2019, 78(FEB.):1–11. KonecnyJ Note on m-polar fuzzy graph representation of concept lattice: How to really compute m-polar fuzzy concepts [J] Engineering Applications of Artificial Intelligence 2019 78 FEB. 1 11 10.1016/j.engappai.2018.10.015 Search in Google Scholar

Zou L, Feng K, Liu X. Linguistic-Valued Intuitionistic Fuzzy Concept Lattice and Its Application [J]. Journal of Computer Research and Development, 2018. ZouL FengK LiuX Linguistic-Valued Intuitionistic Fuzzy Concept Lattice and Its Application [J] Journal of Computer Research and Development 2018 Search in Google Scholar

Codara P, Esteva F, Godo L, et al. Connecting Systems of Mathematical Fuzzy Logic with Fuzzy Concept Lattices [C]//International Conference on Information Processing and Management of Uncertainty in Knowledge-Based Systems. Springer, Cham, 2018. CodaraP EstevaF GodoL Connecting Systems of Mathematical Fuzzy Logic with Fuzzy Concept Lattices [C] International Conference on Information Processing and Management of Uncertainty in Knowledge-Based Systems Springer, Cham 2018 10.1007/978-3-319-91476-3_23 Search in Google Scholar

Zhang B, Li C C, Dong Y, et al. A Comparative Study Between Analytic Hierarchy Process and Its Fuzzy Variants: A Perspective based on Two Linguistic Models [J]. IEEE Transactions on Fuzzy Systems, 2020, PP(99):1–1. ZhangB LiC C DongY A Comparative Study Between Analytic Hierarchy Process and Its Fuzzy Variants: A Perspective based on Two Linguistic Models [J] IEEE Transactions on Fuzzy Systems 2020 PP(99):1–1. 10.1109/TFUZZ.2020.3018110 Search in Google Scholar

Mao H, Zheng Z. The construction of fuzzy concept lattice based on weighted complete graph [J]. Journal of Intelligent and Fuzzy Systems, 2019, 36(2):1–9. MaoH ZhengZ The construction of fuzzy concept lattice based on weighted complete graph [J] Journal of Intelligent and Fuzzy Systems 2019 36 2 1 9 10.3233/JIFS-181642 Search in Google Scholar

Li Z A, Hl B, Xs C, et al. Rule extraction based on linguistic-valued intuitionistic fuzzy layered concept lattice [J]. International Journal of Approximate Reasoning, 2021. LiZ A HlB XsC Rule extraction based on linguistic-valued intuitionistic fuzzy layered concept lattice [J] International Journal of Approximate Reasoning 2021 Search in Google Scholar

Li Y, Cao Y, Hui C, et al. A rule extraction method based on intuitionistic fuzzy linguistic concept lattice [J]. Journal of Nanjing University (Natural Science), 2018. LiY CaoY HuiC A rule extraction method based on intuitionistic fuzzy linguistic concept lattice [J] Journal of Nanjing University (Natural Science) 2018 Search in Google Scholar

Zhang Y, Huang D, Lin H, et al. Knowledge reasoning approach with linguistic-valued intuitionistic fuzzy credibility [J]. International Journal of Machine Learning and Cybernetics, 2019, 11(13). ZhangY HuangD LinH Knowledge reasoning approach with linguistic-valued intuitionistic fuzzy credibility [J] International Journal of Machine Learning and Cybernetics 2019 11 13 10.1007/s13042-019-00965-y Search in Google Scholar

Liu P, Liu X. Linguistic Intuitionistic Fuzzy Hamy Mean Operators and Their Application to Multiple-Attribute Group Decision Making [J]. IEEE Access, 2019, 7:127728–127744. LiuP LiuX Linguistic Intuitionistic Fuzzy Hamy Mean Operators and Their Application to Multiple-Attribute Group Decision Making [J] IEEE Access 2019 7 127728 127744 10.1109/ACCESS.2019.2937854 Search in Google Scholar

Meng F, Jie T, Hamido F. Linguistic intuitionistic fuzzy preference relations and their application to multi-criteria decision making [J]. Information Fusion, 2018, 46:77–90. MengF JieT HamidoF Linguistic intuitionistic fuzzy preference relations and their application to multi-criteria decision making [J] Information Fusion 2018 46 77 90 10.1016/j.inffus.2018.05.001 Search in Google Scholar

Fma B, Jie T B, Yz B. Programming model-based group decision making with multiplicative linguistic intuitionistic fuzzy preference relations [J]. Computers & Industrial Engineering, 2019, 136:212–224. FmaB JieT B YzB Programming model-based group decision making with multiplicative linguistic intuitionistic fuzzy preference relations [J] Computers & Industrial Engineering 2019 136 212 224 10.1016/j.cie.2019.07.019 Search in Google Scholar

Liu P, Liu X, Ma G, et al. A Multi-Attribute Group Decision-Making Method Based on Linguistic Intuitionistic Fuzzy Numbers and Dempster–Shafer Evidence Theory [J]. International Journal of Information Technology and Decision Making, 2020, 19(5). LiuP LiuX MaG A Multi-Attribute Group Decision-Making Method Based on Linguistic Intuitionistic Fuzzy Numbers and Dempster–Shafer Evidence Theory [J] International Journal of Information Technology and Decision Making 2020 19 5 10.1142/S0219622020500042 Search in Google Scholar

Bao J, Wang X, Yu Z, et al. Risk Assessment for Ships Based on Information Entropy and Intuitionistic Fuzzy Linguistic Variables [C]//20th COTA International Conference of Transportation Professionals. 2020. BaoJ WangX YuZ Risk Assessment for Ships Based on Information Entropy and Intuitionistic Fuzzy Linguistic Variables [C] 20th COTA International Conference of Transportation Professionals 2020 10.1061/9780784483053.234 Search in Google Scholar

J Tang, Meng F. Linguistic intuitionistic fuzzy Hamacher aggregation operators and their application to group decision making [J]. Granular Computing, 2019, 4(1):109–124. TangJ MengF Linguistic intuitionistic fuzzy Hamacher aggregation operators and their application to group decision making [J] Granular Computing 2019 4 1 109 124 10.1007/s41066-018-0089-2 Search in Google Scholar

Wang X K, Peng H G, Wang J Q. Hesitant linguistic intuitionistic fuzzy sets and their application in multicriteria decision-making problems [J]. International journal for uncertainty quantifications, 2018, 8(4):321–341. WangX K PengH G WangJ Q Hesitant linguistic intuitionistic fuzzy sets and their application in multicriteria decision-making problems [J] International journal for uncertainty quantifications 2018 8 4 321 341 10.1615/Int.J.UncertaintyQuantification.2018019996 Search in Google Scholar

Jie T, Meng F. Linguistic intuitionistic fuzzy Hamacher aggregation operators and their application to group decision making [J]. Granular Computing, 2018(5):1–16. JieT MengF Linguistic intuitionistic fuzzy Hamacher aggregation operators and their application to group decision making [J] Granular Computing 2018 5 1 16 Search in Google Scholar

Teng F, Liu P. Multiple-Attribute Group Decision-Making Method Based on the Linguistic Intuitionistic Fuzzy Density Hybrid Weighted Averaging Operator [J]. International Journal of Fuzzy Systems, 2019. TengF LiuP Multiple-Attribute Group Decision-Making Method Based on the Linguistic Intuitionistic Fuzzy Density Hybrid Weighted Averaging Operator [J] International Journal of Fuzzy Systems 2019 10.1007/s40815-018-0526-z Search in Google Scholar

Liu P, Liu J. A Multiple Attribute Group Decision-making Method Based on the Partitioned Bonferroni Mean of Linguistic Intuitionistic Fuzzy Numbers [J]. Cognitive Computation, 2020, 12(1):49–70. LiuP LiuJ A Multiple Attribute Group Decision-making Method Based on the Partitioned Bonferroni Mean of Linguistic Intuitionistic Fuzzy Numbers [J] Cognitive Computation 2020 12 1 49 70 10.1007/s12559-019-09676-6 Search in Google Scholar

Liu P, Liu W. Scaled Prioritized Operators Based on the Linguistic Intuitionistic Fuzzy Numbers and Their Applications to Multi-Attribute Decision Making [J]. International Journal of Fuzzy Systems, 2018, 20(5):1539–1550. LiuP LiuW Scaled Prioritized Operators Based on the Linguistic Intuitionistic Fuzzy Numbers and Their Applications to Multi-Attribute Decision Making [J] International Journal of Fuzzy Systems 2018 20 5 1539 1550 10.1007/s40815-018-0459-6 Search in Google Scholar

Bhaumik A, Roy S K, Weber G W. Hesitant interval-valued intuitionistic fuzzy-linguistic term set approach in Prisoners’ dilemma game theory using TOPSIS: a case study on Human-trafficking [J]. Central European Journal of Operations Research, 2020, 28(2):797–816. BhaumikA RoyS K WeberG W Hesitant interval-valued intuitionistic fuzzy-linguistic term set approach in Prisoners’ dilemma game theory using TOPSIS: a case study on Human-trafficking [J] Central European Journal of Operations Research 2020 28 2 797 816 10.1007/s10100-019-00638-9 Search in Google Scholar

Liu P, Shen M. An extended C-TODIM method with linguistic intuitionistic fuzzy numbers [J]. Journal of Intelligent and Fuzzy Systems, 2019, 37(1):1–13. LiuP ShenM An extended C-TODIM method with linguistic intuitionistic fuzzy numbers [J] Journal of Intelligent and Fuzzy Systems 2019 37 1 1 13 10.3233/JIFS-182554 Search in Google Scholar

Chu Y, Liu P, Li H. Multi-attribute group decision making method based on some trapezoid intuitionistic fuzzy linguistic Bonferroni mean aggregation operators [J]. Journal of Intelligent and Fuzzy Systems, 2019, 36(4):3869–3889. ChuY LiuP LiH Multi-attribute group decision making method based on some trapezoid intuitionistic fuzzy linguistic Bonferroni mean aggregation operators [J] Journal of Intelligent and Fuzzy Systems 2019 36 4 3869 3889 10.3233/JIFS-181045 Search in Google Scholar

Liu P, You X. Some linguistic intuitionistic fuzzy Heronian mean operators based on Einstein T-norm and T-conorm and their application to decision-making [J]. Journal of Intelligent and Fuzzy Systems, 2018, 35:1–13. LiuP YouX Some linguistic intuitionistic fuzzy Heronian mean operators based on Einstein T-norm and T-conorm and their application to decision-making [J] Journal of Intelligent and Fuzzy Systems 2018 35 1 13 10.3233/JIFS-18032 Search in Google Scholar

Yuan R, Yi L, Zheng P. Novel Multiple Attribute Group Decision-Making Methods Based on Linguistic Intuitionistic Fuzzy Information [J]. 2020. YuanR YiL ZhengP Novel Multiple Attribute Group Decision-Making Methods Based on Linguistic Intuitionistic Fuzzy Information [J] 2020 Search in Google Scholar

Bhaumik A, Roy S K, Weber G W. Hesitant interval-valued intuitionistic fuzzy-linguistic term set approach in Prisoners’ dilemma game theory using TOPSIS: a case study on Human-trafficking [J]. Central European Journal of Operations Research, 2020, 28. BhaumikA RoyS K WeberG W Hesitant interval-valued intuitionistic fuzzy-linguistic term set approach in Prisoners’ dilemma game theory using TOPSIS: a case study on Human-trafficking [J] Central European Journal of Operations Research 2020 28 10.1007/s10100-019-00638-9 Search in Google Scholar

Recommended articles from Trend MD

Plan your remote conference with Sciendo