# Personalized individual semantics-based approach for linguistic failure modes and effects analysis with incomplete preference information

25 Feb 2020-Vol. 52, Iss: 11, pp 1275-1296

TL;DR: This article presents the design of a PIS-based FMEA approach, in which members express their opinions over failure modes and risk factors using Linguistic Distribution Assessment Matrices (LDAMs) and also provide their opinionsover failure modes using incomplete Additive Preference Relations (APRs).

Abstract: Failure Modes and Effects Analysis (FMEA) is a very useful reliability-management instrument for detecting and mitigating risks in various fields. The linguistic assessment approach has recently be...

## Summary (1 min read)

Jump to: [IISE Transactions -For Peer Review] – [(3) Numerical scale function] – [5. Case study] and [2) Comparison with random data]

### IISE Transactions -For Peer Review

- 2 The RPN-based FMEA approach has been associated with several issues (see [1, 2, 10, 11, 23, 37, 38, 41, 46] ), being particularly relevant to the present paper that FMEA members are obliged to express accurate risk assessment information using the aforementioned 1-10 numerical points scale.
- Indeed, in some real-world decision processes, FMEA members may prefer or feel more comfortable assessing risk using linguistic rather than numerical values (e.g., [16, 21, 40] ).
- IISE Transactions -For Peer Review 3 (APRs) on the failure modes.
- Section 2 includes the necessary preliminary concepts to make this paper self-contained.
- It should be noted that the PIS-based linguistic FMEA approach is still useful with the setting of different parameter values to the above ones.

### (3) Numerical scale function

- The concept of numerical scale function was proposed to transform linguistic terms into real numbers [7] , with the aim to facilitate the computational process in the linguistic assessment approach based GDM.
- IISE Transactions -For Peer Review linguistic GDM, and the GDM problem is referred to as a PIS-based linguistic GDM [17] [18] [19] .

### 5. Case study

- This section shows the practical use of the PIS-based linguistic FMEA approach to the problem of the reliability management of blood transfusion [25, 27] .
- For the sake of clarity and readability.

### 2) Comparison with random data

- In order to obtain compelling results, Simulation methods I and II with randomly generated data are designed to compare the PIS and FNS based linguistic FMEA approaches.
- The basic idea of Simulation methods I and II (see approach are smaller than those under the FNS-based linguistic FMEA approach in the three parameter scenarios, which is again consistent with the results obtained using the case study data.
- These findings show that taking the PIS issue into account the PIS-based linguistic FMEA approach can improve the reliability management quality.
- (2) Psychological behaviors, such as non-cooperative behaviors [35] and prospect theory [41] , of FMEA members play an important role in practical FMEA problems.

Did you find this useful? Give us your feedback

1

Personalized individual semantics-based approach for

linguistic

failure modes and effects analysis with incomplete

preference information

Abstract: Failure modes and effects analysis (FMEA) is a very useful reliability-management

instrument for detecting and mitigating risks in various fields. Linguistic assessment approach has

recently been widely used in FMEA. Words mean different things to different people, so FMEA

members may present personalized individual semantics (PIS) in their linguistic assessment

information. This paper designs a PIS-based FMEA approach with members expressing their

opinions over failure modes and risk factors using linguistic distribution assessment matrices

(LDAMs) and also provide their opinions over failure modes using incomplete additive preference

relations (APRs).

A preference information preprocessing method with a two-stage optimization

model is presented to generate complete APRs with acceptable consistency levels from incomplete

APRs. Then, a deviation minimum-based optimization model is designed to personalize individual

semantics by minimizing the deviation between APR and the numerical assessment matrix derived

from the corresponding LDAM. This is followed by the developing of a ranking process to

generate the risk ordering of failure modes. A case study and a detailed comparison analysis are

presented to show the effectiveness of the PIS-based linguistic FMEA approach.

Keywords: Reliability management; failure modes and effects analysis; personalized individual

semantics; consistency; optimization

1. Introduction

Failure modes and effects analysis (FMEA) is a very powerful reliability-management

instrument, which is frequently used in product design to identify the most critical causes of a

product’s failure and to mitigate their risks [3, 9, 12, 15, 29, 32, 36]. Thus, risk assessment and

prioritization of failure modes are key issues in FMEA [4, 5, 12, 40]. Traditionally, the risk

assessment information on each failure mode with respect to the three risk factors of occurrence

(O), severity (S), and detection (D) is measured using a (1-10) numerical points scale. The product

of the (O, S, D) numerical risks values is defined as the risk priority number (RPN) of a failure

mode [39], which is subsequently used to produce a risk ordering of failure modes. Different

levels of security control measures for failure modes are implemented in order to mitigate risk,

and failure modes with high RPN values are paid more attention. Thus, the ultimate decision result

in an FMEA method/model is the risk ordering of failure modes by their RPN values. For a

comprehensive introduction of the FMEA implementation process, please refer to Refs. [22, 39].

Page 6 of 39

For ScholarOne support, you may contact 434-964-4100 or 888-503-1050 (US-based numbers).

IISE Transactions - For Peer Review

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57

58

59

60

2

The RPN-based FMEA approach has been associated with several issues (see [1, 2, 10, 11, 23,

37, 38, 41, 46]), being particularly relevant to the present paper that FMEA members are obliged

to express accurate risk assessment information using the aforementioned 1-10 numerical points

scale. Indeed, in some real-world decision processes, FMEA members may prefer or feel more

comfortable assessing risk using linguistic rather than numerical values (e.g., [16, 21, 40]). To

address this issue, the following different linguistic FMEA approaches have been

developed/proposed to date: (i) interval two-tuple linguistic risk assessments and the ELECTRE

(Elimination Et Choix Traduisant la REalite) approach for ranking failure modes [24]; (ii)

triangular fuzzy linguistic risk assessments, the Choquet integral and prospect theory [41]; (iii) the

linguistic weighted geometric operator and fuzzy priority methodology [47]; (iv) linguistic

distribution of risk assessment information with an improved TODIM (Portuguese acronym for

‘interactive and multi-criteria decision making’) approach to yield the risk ordering of failure

modes [14]; (v) multi-granular linguistic distribution risk assessments to model uncertain opinions

of FMEA members [34]; (vi) consensus-based group decision-making (GDM) approaches for

FMEA with linguistic distribution risk assessments [44, 45].

These linguistic FMEA approaches represent a great progress because they provide a more

flexible framework than the previous numerical approaches. However, they still do not

accommodate all possible real scenarios because they are based on the premise that the linguistic

labels (words) implemented/used mean the same for all FMEA members, when the reality is that,

in general, words may mean different things to different individuals, a phenomenon referred to as

personalized individual semantics (PIS) that affects practical FMEA problems decision results

([17-19, 30, 31]). Indeed, when assessing the risk level of a failure mode, two FMEA members

may assess the risk level of the failure mode with the same word “high” but with distinct

semantics. As it is illustrated later in this paper in Section 3.1 (Example 1), ignoring the PIS issue

affects the reliability management quality.

To the best of our knowledge, PIS has not been considered yet in the existent linguistic FMEA

approaches, which is the goal of this paper. Thus, this paper aims at developing the mathematical

framework for managing PIS in the linguistic FMEA problem to improve the reliability

management quality. This is achieved by means of the following two main research objectives:

(1) To implement a general linguistic decision context to formulate the PIS-based FMEA

problem, which is based on the mathematical representation of the FMEA members’ preference

information using (i) the general linguistic model of linguistic distribution assessment matrices

(LDAMs) [6, 14, 20, 28, 42, 43] for the risk assessment information on the failure modes with

respect to the risk factors (O, S, D); and (ii) general incomplete additive preference relations

Page 7 of 39

For ScholarOne support, you may contact 434-964-4100 or 888-503-1050 (US-based numbers).

IISE Transactions - For Peer Review

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57

58

59

60

3

(APRs) on the failure modes.

(2) To obtain the resolution procedure to deal with the PIS-based linguistic FMEA problem: (i)

First, a preference information preprocessing method based on a two-stage optimization model is

constructed to generate complete APRs with acceptable consistency levels from the incomplete

APRs; (ii) Second, inspired by the numerical scale function-based approaches for dealing with PIS

reported in [17-19], a deviation minimum-based optimization model is proposed to personalize

individual semantics in the LADMs, which seeks to minimize the deviation between the APR and

the numerical assessment matrix (NAM) derived from the LDAM using a PIS-based numerical

scale (PNS); (iii) Third, a ranking process is presented to generate the risk ordering of failure

modes from the obtained NAMs and APRs.

Finally, a validation study is reported with a detailed comparison analysis between the proposed

PIS-based linguistic FMEA approach and the fixed numerical scale (FNS) based linguistic FMEA

approach, which is complemented with the exemplification of its practical use with a case study

related to the problem of the reliability management of blood transfusion.

The remainder of this paper is arranged as follows. Section 2 includes the necessary preliminary

concepts to make this paper self-contained. Section 3 presents a motivation example and

formulates the PIS-based linguistic FMEA problem. Section 4 designs the detailed solution

procedure for the PIS-based linguistic FMEA problem. Following this, Section 5 illustrates the

practical use of the PIS-based linguistic FMEA approach with the above mentioned case study.

Subsequently, Section 6 presents a comparison analysis to show the effectiveness of the PIS-based

linguistic FMEA approach. Finally, Section 7 concludes the paper and discusses future research

directions.

In order to improve readability, all acronyms and notations used within the paper are included in

Appendix A.

2. Preliminaries

knowledge regarding APRs, the linguistic distribution assessments,

This section introduces basic

and the numerical scale function of a linguistic term set.

(1) Additive preference relations (APRs)

Let

12

{ , , ..., }

n

X x x x

be a set of objects. The concept of APR over

X

and its consistency

level are provided in the below definitions:

Definition 1 [13]. An APR on a set of objects

X

is represented by a matrix,

()

ij n n

Aa

, in

which its element

[0,1]

ij

a

represents the preference intensity of object

i

x

over object

j

x

,

subject to the following reciprocity property:

1

ij ji

aa

, {1,2,..., }i j n

.

Definition 2 [13]. The consistency level of an APR

()

ij n n

Aa

is measured with the

Page 8 of 39

For ScholarOne support, you may contact 434-964-4100 or 888-503-1050 (US-based numbers).

IISE Transactions - For Peer Review

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57

58

59

60

4

following [0,1]-valued function:

, , 1;

4

( ) 1 | 0.5|

( 1)( 2)

n

ij jz iz

i j z i j z

CL A a a a

n n n

. (1)

The larger the value of

()CL A

, the more consistent the APR

()

ij n n

Aa

is. In particular,

()

ij n n

Aa

is completely consistent when

( ) 1CL A

. Let

[0,1]

be a predefined threshold

value used to judge whether the consistency of

()

ij n n

Aa

is acceptable or not, i.e. when

()CL A

, then APR

()

ij n n

Aa

is of acceptable consistency; otherwise, it is not of acceptable

consistency. The determination of the parameter

usually depends on the actual

situation/problem being dealt with.

The complexity of a particular decision context may lead to a decision maker not been able to

()

ij n n

Aa

. In this case, the APR is referred to as an incomplete

provide all elements of an APR

APR. We use

()

ij n n

Aa

to denote an incomplete APR. In particular,

ij

a null

if the preference

intensity of object

i

x

over object

j

x

is not provided by the decision maker.

(2) Linguistic distribution assessments

Let

0

{ , ..., }

g

L l l

be a linguistic term set with granularity of

1g

and term

j

l

a possible

linguistic value subject to the following two conditions: (1)

L

is ordered:

ij

l l i j

, and (2)

there is an inverse function such that

()

j g j

neg l l

. The linguistic distribution assessment

approach was proposed to deal with uncertain and vague assessment information effectively [43].

The linguistic distribution assessment is formally presented below.

Definition 3 [43]. A distribution assessment of a linguistic term set

L

is represented as

{( , )| 0,1,..., }

tt

LAD l t g

, with symbolic proportions

[0,1]

t

of linguistic terms

t

l

satisfying

0

1

g

t

t

.

(3) Numerical scale function

The concept of numerical scale function was proposed to transform linguistic terms into real

numbers [7], with the aim to facilitate the computational process in the linguistic assessment

approach based GDM.

Definition 4. Let

0

{ , ..., }

g

L l l

be a linguistic term set and

RN

be the set of real numbers.

A function

:NS L RN

is called a numerical scale of

L

, and

()

i

NS l

is the numerical index of

i

l

. If the function

NS

is strictly monotone increasing, then

NS

is called an ordered numerical

scale of

L

.

In essence, the numerical scale provides numerical meaning of linguistic terms in GDM. When

all individuals in a GDM problem have the same numerical indexes of linguistic terms, then the

GDM problem is referred to as an FNS-based linguistic GDM. In many cases though, words mean

different things to different individuals, i.e. individuals have different PIS, PNSs are employed in

Page 9 of 39

For ScholarOne support, you may contact 434-964-4100 or 888-503-1050 (US-based numbers).

IISE Transactions - For Peer Review

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57

58

59

60

5

linguistic GDM, and the GDM problem is referred to as a PIS-based linguistic GDM [17-19].

3. The PIS-based linguistic FMEA problem: Motivation and problem formulation

This section includes a motivation example to analyze the influence of PIS on linguistic FMEA.

Then, the PIS-based linguistic FMEA problem formulation is put forward.

3.1. Motivation example: The influence of PIS on linguistic FMEA

It was argued in the introduction section that the PIS issue may appear in the linguistic

assessments of the FMEA members and that ignoring it weakens the reliability management

quality. Here, we provide an example that provides evidence to support this argument.

Example 1: Let

0

{ very lowLl

,

1

lowl

,

2

moderately lowl

,

3

moderatel

,

4

moderate highl

,

5

highl

,

6

very high}l

be the linguistic term set used by FMEA members

1

TM

and

2

TM

to evaluate the risk level of failure modes

1

FM

and

2

FM

. The aim is to find a

risk ordering of

1

FM

and

2

FM

based on the linguistic risk assessments provided by

1

TM

and

2

TM

, who considered equally important. Let us assume that the risk assessments on

12

( , )FM FM

provided by

1

TM

and

2

TM

are (

5

l

and

4

l

) and (

3

l

and

5

l

), respectively.

The process to obtain the risk ordering of

1

FM

and

2

FM

is as follows: (1) linguistic

assessments are transformed into numerical assessments using a numerical scale function as per

Definition 4; (2) the total evaluation value (TEV) of each failure mode is obtained as the weighted

average of the numerical assessments of the two FMEA members, which in this case will be the

mean value as both members are equally important; (3) the TEV values are be used to generate the

risk ordering of

1

FM

and

2

FM

.

In the following, two cases are considered:

Case A: PIS is not considered and

1

TM

and

2

TM

use the same following numerical scale

function:

12

( ) ( ) / 6

ii

NS l NS l i

( 0,1,...,6)i

. Thus, it is

12

1 5 3

0.5( ( ) ( ))TEV NS l NS l

0.5(5/6 0.5) 0.6667

and

12

2 4 5

0.5( ( ) ( )) 0.5(2 / 3 5 / 6) 0.75TEV NS l NS l

.

Case B: PIS is considered and

1

TM

and

2

TM

use different numerical scale functions:

1,* 1,*

06

{ ( ),..., ( )}={0, 1/6, 1/3, 1/2, 0.52, 0.92, 1}NS l NS l

and

2,* 2,*

06

{ ( ),..., ( )} {0 / 6, 1/ 6, 1/ 3, 1/2,NS l NS l

2/3, 5/6, 6/6}

, respectively. Thus, it is

* 1,* 2,*

1 5 3

0.5( ( ) ( )) 0.5(0.92 0.5) 0.71TEV NS l NS l

and

* 1,*

24

0.5( ( )TEV NS l

2,*

5

( ))NS l

0.5(0.52 5/ 6) 0.6767

.

In case A, the risk level of

2

TM

is higher than

1

TM

. However, the risk level of

1

TM

is

higher than

2

TM

in case B. This example shows that ignoring the PIS issue may result in

obtaining the opposite risk ordering of failure modes. So, it is worth to design an approach to

address the PIS issue in the linguistic FMEA problem when it is present.

3.2. Problem formulation

In the following, we formulate the PIS-based FMEA problem in a general linguistic decision

Page 10 of 39

For ScholarOne support, you may contact 434-964-4100 or 888-503-1050 (US-based numbers).

IISE Transactions - For Peer Review

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57

58

59

60

##### Citations

More filters

••

TL;DR: This work presents the taxonomy of existing distributed linguistic representations, and reviews the key elements and applications of distributed linguistic information processing in decision making, including the distance measurement, aggregation methods, distributed linguistic preference relations, and distributed linguistic multiple attribute decision making models.

123 citations

••

TL;DR: In this article, a simplified linguistic computational model is proposed to fuse multi-granular unbalanced linguistic terms for multi-criteria group decision making problems and two optimization models are developed to generate adjustment advice for decision makers who have to change their opinions in consensus reaching process, which consider both the bounded confidence levels and minimum adjustment of decision makers' linguistic assessments.

118 citations

••

TL;DR: An approach to TSMDM with multi-granular HFLTSs is developed and allows matching objects to provide linguistic assessments flexibly and can deal with the situations when incomplete criteria weight information is provided.

Abstract: Two-sided matching decision making (TSMDM) problems exist widely in human being’s daily life. For practical TSMDM problems, matching objects with different culture and knowledge backgrounds usually tend to provide linguistic assessments using different linguistic term sets (i.e., multi-granular linguistic information). Moreover, for TSMDM problems with high uncertainty, it is possible that matching objects may have some hesitancy and thus provide hesitant fuzzy linguistic term sets (HFLTSs). To model these situations, an approach to TSMDM with multi-granular HFLTSs is developed in the paper. In the proposed approach, some optimization models are first constructed to determine criteria weights for matching objects who do not provide clear criteria weight vectors. Afterwards, each matching object’s hesitant fuzzy linguistic decision matrix is aggregated to obtain his/her collective assessments over matching objects on the other side, which are denoted by multi-granular linguistic distribution assessments. These multi-granular linguistic distribution assessments are unified to obtain matching objects’ satisfaction degrees. Furthermore, an optimization model which aims to maximize the overall satisfaction degree of matching objects by considering the stable matching condition is then established and solved to determine the matching between matching objects. Eventually, an example for the matching of green building technology supply and demand is provided to demonstrate the characteristics of the proposed approach. Compared with previous studies, the proposed approach allows matching objects to provide linguistic assessments flexibly and can deal with the situations when incomplete criteria weight information is provided.

89 citations

••

TL;DR: In this article , a PIS-based individual consensus-level maximization model and a minimum adjustment-based optimization model for linguistic group decision making (GDM) is proposed.

Abstract: Consistency and consensus are important issues for linguistic group decision making (GDM), which have been extensively studied by scholars. Nevertheless, most of previous consensus reaching models focus on adjusting decision makers’ preference relations and ignore the individual consistency, which results in that individual consistency may be destroyed by using these consensus reaching models. Moreover, it has been accepted that words mean different things for different people and thus, it is also necessary to model decision makers’ personalized individual semantics (PISs) in linguistic GDM. This work focuses on developing some PIS-based consistency control and consensus reaching models for linguistic GDM. First, we analyze the problems existing in previous PIS models and then develop a minimum adjustment-based optimization model to test and improve the individual consistency for a linguistic preference relation (LPR). Followed by this, a PIS-based individual consensus-level maximization model and a PIS-based minimum adjustment model are established for consensus reaching in linguistic GDM, in which individual consistency control is considered. Furthermore, an algorithm for consensus reaching is proposed based on these models. To justify the proposed models and algorithm, some numerical results and simulation analysis are provided eventually.

56 citations

##### References

More filters

••

TL;DR: In this paper, failure mode and effect analysis: Failure Mode and Effect Analysis: FMEA From Theory to Execution Technometrics: Vol 38, No 1, pp 80-80

Abstract: (1996) Failure Mode and Effect Analysis: FMEA From Theory to Execution Technometrics: Vol 38, No 1, pp 80-80

1,287 citations

### "Personalized individual semantics-b..." refers methods in this paper

...mode [39], which is subsequently used to produce a risk ordering of failure modes....

[...]

••

TL;DR: This study reviewed 75 FMEA papers published between 1992 and 2012 in the international journals and categorized them according to the approaches used to overcome the limitations of the conventional RPN method.

Abstract: Failure mode and effects analysis (FMEA) is a risk assessment tool that mitigates potential failures in systems, processes, designs or services and has been used in a wide range of industries. The conventional risk priority number (RPN) method has been criticized to have many deficiencies and various risk priority models have been proposed in the literature to enhance the performance of FMEA. However, there has been no literature review on this topic. In this study, we reviewed 75 FMEA papers published between 1992 and 2012 in the international journals and categorized them according to the approaches used to overcome the limitations of the conventional RPN method. The intention of this review is to address the following three questions: (i) Which shortcomings attract the most attention? (ii) Which approaches are the most popular? (iii) Is there any inadequacy of the approaches? The answers to these questions will give an indication of current trends in research and the best direction for future research in order to further address the known deficiencies associated with the traditional FMEA.

666 citations

••

01 Feb 2007

TL;DR: This paper proposes an iterative procedure to estimate the missing information in an expert's incomplete fuzzy preference relation, guided by the additive-consistency (AC) property, and proposes a new induced ordered weighted averaging operator, the AC-IOWA operator, which permits the aggregation of the experts' preferences in such a way that more importance is given to the most consistent ones.

Abstract: In decision-making problems there may be cases in which experts do not have an in-depth knowledge of the problem to be solved. In such cases, experts may not put their opinion forward about certain aspects of the problem, and as a result they may present incomplete preferences, i.e., some preference values may not be given or may be missing. In this paper, we present a new model for group decision making in which experts' preferences can be expressed as incomplete fuzzy preference relations. As part of this decision model, we propose an iterative procedure to estimate the missing information in an expert's incomplete fuzzy preference relation. This procedure is guided by the additive-consistency (AC) property and only uses the preference values the expert provides. The AC property is also used to measure the level of consistency of the information provided by the experts and also to propose a new induced ordered weighted averaging (IOWA) operator, the AC-IOWA operator, which permits the aggregation of the experts' preferences in such a way that more importance is given to the most consistent ones. Finally, the selection of the solution set of alternatives according to the fuzzy majority of the experts is based on two quantifier-guided choice degrees: the dominance and the nondominance degree

556 citations

### "Personalized individual semantics-b..." refers methods in this paper

...The consistency level as per Definition 2 [13] is used as the objective function of the following...

[...]

•

26 Apr 2010

TL;DR: Perceptual Computing explains how to implement CWW to aid in the important area of making subjective judgments, using a methodology that propagates random and linguistic uncertainties into the subjective judgment in a way that can be modeled and observed by the judgment maker.

Abstract: Explains for the first time how "computing with words" can aid in making subjective judgments Lotfi Zadeh, the father of fuzzy logic, coined the phrase "computing with words" (CWW) to describe a methodology in which the objects of computation are words and propositions drawn from a natural language. Perceptual Computing explains how to implement CWW to aid in the important area of making subjective judgments, using a methodology that leads to an interactive devicea "Perceptual Computer"that propagates random and linguistic uncertainties into the subjective judgment in a way that can be modeled and observed by the judgment maker. This book focuses on the three components of a Perceptual Computerencoder, CWW engines, and decoderand then provides detailed applications for each. It uses interval type-2 fuzzy sets (IT2 FSs) and fuzzy logic as the mathematical vehicle for perceptual computing, because such fuzzy sets can model first-order linguistic uncertainties whereas the usual kind of fuzzy sets cannot. Drawing upon the work on subjective judgments that Jerry Mendel and his students completed over the past decade, Perceptual Computing shows readers how to: Map word-data with its inherent uncertainties into an IT2 FS that captures these uncertainties Use uncertainty measures to quantify linguistic uncertainties Compare IT2 FSs by using similarity and rank Compute the subsethood of one IT2 FS in another such set Aggregate disparate data, ranging from numbers to uniformly weighted intervals to nonuniformly weighted intervals to words Aggregate multiple-fired IF-THEN rules so that the integrity of word IT2 FS models is preserved Free MATLAB-based software is also available online so readers can apply the methodology of perceptual computing immediately, and even try to improve upon it. Perceptual Computing is an important go-to for researchers and students in the fields of artificial intelligence and fuzzy logic, as well as for operations researchers, decision makers, psychologists, computer scientists, and computational intelligence experts.

435 citations

••

TL;DR: The concept of distribution assessments in a linguistic term set is proposed, and the operational laws of linguistic distribution assessments are studied, to provide a theoretic basis for the application of linguistic distributions assessments in group decision making.

420 citations

### "Personalized individual semantics-b..." refers background in this paper

...(LDAMs) [6, 14, 20, 28, 42, 43] for the risk assessment information on the failure modes with...

[...]

...approach was proposed to deal with uncertain and vague assessment information effectively [43]....

[...]