=Paper=
{{Paper
|id=Vol-2470/p12
|storemode=property
|title=CWW enhanced fuzzy SWOT evaluation for risk analysis and decision making under uncertainty
|pdfUrl=https://ceur-ws.org/Vol-2470/p12.pdf
|volume=Vol-2470
|authors=Žygimantas Meškauskas
|dblpUrl=https://dblp.org/rec/conf/ivus/Meskauskas19
}}
==CWW enhanced fuzzy SWOT evaluation for risk analysis and decision making under uncertainty==
https://ceur-ws.org/Vol-2470/p12.pdf
CWW Enhanced Fuzzy SWOT Evaluation for Risk
Analysis and Decision Making under Uncertainty
Žygimantas Meškauskas
Department of Computer Sciences
Kaunas University of Technology
Kaunas, Lithuania
zygimantas.meskauskas@ktu.lt
Abstract—The SWOT analysis is a method used II. RELATED WORKS
worldwide to assist in the decision making in industrial, and
business management, as well as in banking, military SWOT analysis enhanced by the ‘Computing with
planning operations, and science. Without question, it is seen Words’ methodology is described in [10]. This article
as an obligatory tool on both the governmental level, as well mainly focuses on the use of analysis under uncertainties
the personal. Until now, all data had to be collected from the for experts’ knowledge extraction, and the use of analysis
experts and the decision makers in numerical form, and be results in risk management and decision making. The idea
presented in numerical form. In this paper, we aim to enrich is that risk is not simply a loss multiplied by the probability,
the SWOT analysis using the ‘Computing with Words’ but that there are also positive risk options, described in [4].
paradigm for expert knowledge extraction in a verbal form. The risk management part in this work is based on a
By presenting data in this format, we allow experts to express
composed risk formula, presented in [7], that links risk
their opinion alongside possible uncertainties. Moreover,
enriched SWOT analysis results are extremely useful for the analysis inputs and SWOT analysis outputs.
risk analysis and decision making.
III. CWW ENHANCED SWOT ANALYSIS
Keywords—SWOT analysis, computing with words, fuzzy, It is known that SWOT stands for strengths (ST),
risk analysis, decision making, uncertainty weaknesses (WK), opportunities (OP), and threats (TH) that
surround any idea, plan, or project to be investigated
I. INTRODUCTION and / or implemented. Opportunities and threats are usually
There are many numerous methods for extracting defined as external issues of the project and signify possible
knowledge from experts throughout the varying fields of positive and negative achievements once the project is
academic and professional activity. If some information realized. At the same time, strengths and weaknesses mean
about one specific area is needed, it is not mandatory to internal issues enable, and impede, the achievement of both
have deep knowledge in that area. This is the case where the main goals and the development of projects. A
field experts take a major role, and the method itself is only quantitative interaction between OPs, THs, STs and WKs is
needed to save extracted information in a structured form. usually expressed by a numerical SWOT matrix which
Generally, data extraction and the structuring process can shows the influence of STs and WKs on strengths and
be defined as: threats [10].
This article aims to find ways on how to use verbal
Data → Information → Knowledge → Wisdom. qualitative evaluation in the process of delivering
Data extraction is always performed in a certain form of descriptions of data necessary for SWOT analysis.
dialogue. Experts from different fields often use different Attempting to perform necessary SWOT computations and
deliver the obtained SWOT analysis results in a verbal form
terminology to describe the same objects, just from
OPs, THs, STs and WKs were characterized by means of
different perspectives. The biggest challenge is to conduct a using words. It indicates that CWW (Computing with
successful conversation with an expert so that the opinion Words) methodology enriches SWOT methodology and
would be expressed adequately. For this purpose, a widely creates a possibility for SWOT users and decision makers
used SWOT analysis method, enriched with the ‘computing to communicate using words of common language. We
with words’ paradigm, was used for a verbal knowledge propose and investigate new possibilities to apply and
expression and uncertainties evaluation. The results of such enrich SWOT analysis mechanisms, using elements of
analysis can also be expressed in linguistic form, providing artificial intelligence, and the computing with words
information for the risk management and decision making. paradigm. This approach is novel due to the originality of
the encoding of input words that describe the investigated
Chapter 2 contains a related work section, chapter 3 situation in a new functional organization of the SWOT
describes CWW enhanced SWOT analysis methodology, engines. Put simply, the method, decodes and aggregates
and chapter 4 describes risk management and decision numerical outputs into a verbal form. The main idea of
making. In chapter 5, experimental simulation is presented, CWW enhanced SWOT analysis is to take verbal
and chapter 6 concludes everything with remarks. descriptions as input, convert that data into numbers for
internal computation using a ‘fuzzy logic’ engine, and
translate the result to the user in a verbal form (as shown in
Fig. 1).
© 2019 for this paper by its authors. Use permitted under Creative
Commons License Attribution 4.0 International (CC BY 4.0)
38
Fig. 1. Functional structure of SWOT+CWW methodology.
It is not necessary to have the knowledge on a specific (µ) as 0.8. Left shoulder of term “Large” (XL(L)) is the
domain — this is the role of experts. A certain number of pessimistic value of uncertainty and the right shoulder
experts can describe the situation and all the dynamics of (XL(R)) is optimistic.
the domain. The main focus is to collect required expert
When all data needed for SWOT analysis is submitted
information for analysis and data storing. The most
in that form, aggregated opportunity and values
convenient way to describe a situation for any human being
is to express it verbally instead of using numbers but some are calculated. Due to the data being translated in two ways
level of uncertainty arises from those words. Computational (pessimistic and optimistic), there is a possibility for
systems are based on a numerical data, so data encoding, multiple perspectives of the results that can serve as a
and decoding, is needed. In line with the CWW paradigm, possible input data for risk analysis methodology.
all inputs and outputs to the user (expert) are in a verbal IV. RISK ANALYSIS AND DECISION MAKING
form. All the internal SWOT analysis computations using
CWW paradigm are performed using a black box principle. Risk is the level of uncertainty of action (results). Most
When an expert characterizes the information and dynamics of the methodologies interpret that risk directly depends on
for the domain, all this information and data is processed by threats. In our approach we reference to Hillson [4] and
a translated list of rules and algorithms. Rules and state that risk is symbiosis of opportunities and threats. To
algorithms are determined by expert’s described dynamics implement this idea, we have associated risk components
of the field and used to translate between numerical and with SWOT analysis.
verbal data using ‘Fuzzification’ and ‘Defuzzification’ with A. Risk analysis
fixed membership functions (displayed in Fig. 2).
In the context of a risk analysis, opportunities and
threats can be associated with SWOT analysis components
with opportunities and threats components; efforts and
hesitancies also make an impact. Efforts can be expressed
as investments in a risk analysis process, and hesitancies are
the level of uncertainty. In our approach, risk can be
described as:
R = R(EFF↑; OP↓; TH↑; HES ↑)
Fig. 2. Fixed CWW fuzzy membership functions.
The ‘Fuzzy logic’ engine calculates a numerical value The concept of risk combines:
of a given verbal term and a value of uncertainty by Activity (EFF/efforts/input/ ...);
assigning a membership function. The number of different
verbal terms describes input words as possibilities. But, Potentially positive results
according to “Miller’s law” [6] (The Magical Number (OP/ achievements/attainments/ ...);
Seven, Plus or Minus Two), a human can differentiate
approximately up to seven different verbal evaluations. This Potentially negative results (TH/ losses/defeats, …);
CWW enhanced SWOT analysis verbal data input and Uncertainties
output dictionary are selected based on this law. It used six (HES/hesitations/instabilities/options/probabilities/
different terms:
...).
“Zero” ({Z}), OP and TH components of risk are strictly related to
“Very small” ({VS}), SWOT analysis outcomes ( and ). Risk can be
“Small” ({S}), evaluated by combining it with an expert evaluation about
“Medium” ({M}), required efforts (EFF) and (if needed) uncertainties (HES)
“Large” ({L}), evaluation. Risk evaluation can be estimated, and actions
“Very large” ({VL}). taken if necessary. Furthermore, verbal advices or visual
Each verbal term from the selected dictionary has its representation of the results can be done.
triangular form. The peak of each triangle on the X axis
represents a numerical value for verbal terms in case of an B. Decision making
absolute certainty. Left and right shoulders of the triangle A decision is a commitment to a proposition, or a plan
represent uncertainty. In an example (Fig. 2), an expert of an action based on the information and values associated
expressed an opinion as “Large” with a degree of certainty with the possible outcomes. The process operates in a
39
flexible timeframe that is free from the immediacy of Fig. 3. Opportunity input.
evidence acquisition and the real time demands of the
action itself. Thus, it involves deliberation, planning, and The second step in data input procedure is ST and WK
strategizing [8]. The study of decision making is a information as well as the data of influences. Information
multidisciplinary field. It occurs in psychology, statistics, about strength or weakness is entered analogous to
economics, finance, engineering (e.g., quality control), opportunities and threats. Procedure of the influence input
political science, philosophy, medicine, ethics, and is as follows: the user chooses ST or WK component from
jurisprudence. There are many conflicting criterions that the existing list and then specifies the influenced
need to be evaluated in making decisions in our daily or component (OP or TH). Value of influence is entered in a
professional lives. verbal form. There are three ways to express certainty about
the given evaluation:
Research on a multi-criteria decision support developed
two main groups of methods, i.e., American and European 1. Absolute certainty — used, when there is no doubt
schools. Methods of the American school of decision about given estimate;
support are based on a functional approach, more precisely 2. Digital certainty — used, when there is some
the utility or value function. Researchers from the European uncertainty which can be evaluated;
school emphasize the fact that many methods do not 3. Verbal certainty — possibility to express both
consider the variability and uncertainty of expert evaluation and a level of certainty about that
judgments. However, the most common solution to this evaluation in verbal form.
problem is to use granular mathematics, e.g., fuzzy sets Strength input is shown in Fig. 4.
theory or interval arithmetic [5].
V. EXPERIMENTAL SIMULATION
Generally, a lot of SWOT analysis tools were created,
but they lack verbal operations. For this reason, a
prototypical SWOT enhanced CWW analysis tool was
created and used to test the effectiveness of the described
methodology. Pilot testing was made on “Construction of a
new hotel complex in a particular area” example from [11].
The example itself has already been analyzed in article and
all SWOT analysis data is accessible for the use and the
comparison of the results.
A. Data input
SWOT enhanced CWW tool data input is processed by
one component at a time. There are two groups of identical
data input:
1. Opportunities and Threats; Fig. 4. Strength influence on threat.
2. Strengths and Weaknesses.
The user must enter a title and a short acronym of every B. Testing situation
SWOT analysis component (row number is generated Pilot testing was done using example from [11]. List of
automatically if not specified). When the user submits OP opportunities is shown in the TABLE I.
or TH information, a degree of importance (impact) and
TABLE I. LIST OF OPPORTUNITIES
value of truth (membership value) evaluations needs to be
specified. Estimation itself is entered in a verbal form. The
input of the opportunity is shown in Fig. 3.
List of threats is shown in the TABLE II.
TABLE II. LIST OF THREATS
List of strengths is shown in the TABLE III.
TABLE III. LIST OF STRENGTHS
40
1. Optimistic — the best possible result of an overall
Opportunities and Threats evaluation (Best
opportunities size);
2. Pessimistic — the worst possible result of an overall
Opportunities and Threats evaluation (Worst threats
List of weaknesses is shown in the TABLE IV. size);
TABLE IV. LIST OF WEAKNESSES 3. Medium — the average result of overall Opportunities
and Threats evaluation (Realistic view);
The tool shows numerical results in a graphical form
and verbal results are shown at the bottom as the value and
the certainty. Looking at the pessimistic perspective of this
model, the resulting opportunities are estimated as very
small (VS) with 0.4 certainty and as small (S) with 0.6
All SWOT analysis components and evaluations are
certainty. Meanwhile in the optimistic perspective common
presented in a matrix. A SWOT evaluation matrix is shown
opportunities are estimated as small (S) with 0.83 certainty,
in TABLE V.
and as medium (M) with 0.17 certainty. These results
TABLE V. SWOT EVALUATION MATRIX reflect the hotel complex building in Palanga Lithuania
(example from article [11]).
VI. CONCLUDING REMARKS
This paper suggests the use of verbal descriptions for
SWOT analysis data input. A new prototypical software
tool based on Hillson’s ideology and methodology about
enriching SWOT analysis with the CWW paradigm was
“Degrees of importance” (c), “Values of truth” (ρ) and created. Successful experiment simulation based on a
influences are shown in verbal form (S – small, M- created tool was made and simulation results were
medium, L- large). Some of the words (Z - zero, VS - very presented. Those results can serve as expert information for
small and VL - very large) did not occur in our model. risk management and decision making.
Further research objective is to create a network of tools
C. Experimental results for more complex situation analysis with more than one
The final evaluation of summarized opportunities OP∑ as SWOT analysis possibility. The main idea of SWOT
well as threats TH∑ is performed according to formulas (2) enhanced CWW network is to use one SWOT analysis
and (3): results as an influence on another connected SWOT
analysis results.
(2)
ACKNOWLEDGMENT
(3) I wish to thank prof. Raimundas Jasinevičius for his
SWOT analysis results are shown in Fig. 5. methodological assistance and guidelines and prof. Egidijus
Kazanavičius for creating an environment for the research.
REFERENCES
[1] L. A. Zadeh, „Towards Human Level Machine Intelligence -
Is It Achievable? The Need for Paradigm Shift,“ IEEE Computational
Intelligence Magazine, t. 3, nr. 3, pp. 11-22, September 2008.
[2] S. K. Pal, R. Banerjee, S. Dutta ir S. S. Sarma, „An Insight
Into The Z-number Approach To CWW,“ Fundamenta Informaticae, t.
124, nr. 1-2, pp. 197-229, 2013.
[3] M. J. Kochenderfer, Decision Making Under Uncertainty:
Theory and Application, London: The MIT Press, 2018.
[4] D. Hillson, Effective Opportunity Management for Projects:
Exploiting Positive Risk, New York: Marcel Dekker, Inc., 2004, p. 316.
[5] S. Faizi, T. Rashid, W. Sałabun, S. Zafar and J. Wątróbski,
"Decision Making with Uncertainty Using Hesitant Fuzzy Sets,"
International Journal of Fuzzy Systems, vol. 20, no. 1, pp. 93-103,
January 2018.
[6] G. A. Miller, „The magical number seven, plus or minus two:
some limits on our capacity for processing information,“ Psychological
Review, t. 63, nr. 2, pp. 81-97, 1956.
Fig. 5. Numerical and verbal results. [7] Balžekienė, Aistė; Gaulė, Eglė; Jasinevičius, Raimundas;
Kazanavičius, Egidijus; Petrauskas, Vytautas, „Risk Evaluation: The
By given SWOT analysis evaluations, results are Paradigm and Tools,“ įtraukta Information and Software Technologies:
calculated and presented in three ways: 21st International Conference, ICIST 2015, Druskininkai, Lithuania,
2015.
41
[8] M. N. Shadlen ir R. Kiani, „Decision Making as a Window on [11] Tomasz, K., Nowicki, R.K., and Napoli, C.. "Comparison of
Cognition,“ Neuron, t. 80, nr. 3, pp. 791-806, 30 October 2013. effectiveness of multi-objective genetic algorithms in optimization of
[9] "Artificial Intelligence: How knowledge is created, invertible s-boxes." International Conference on Artificial Intelligence
transferred, and used," Elsevier, 2018. and Soft Computing. Springer, Cham, p. 466-476, 2017.
[10] Petrauskas, Vytautas; Jasinevičius, Raimundas; Kazanavičius, [12] R. Jasinevičius and V. Petrauskas, "Dynamic SWOT Analysis
Egidijus; Meškauskas, Žygimantas;, „CWW elements to enrich SWOT as a Tool for Environmentalists," Environmental Research, Engineering
analysis,“ Journal of Intelligent and Fuzzy Systems, t. 34, nr. 1, pp. 307- & Management, vol. 43, no. 1, pp. 14-20, 2008
320, January 2018.
42