The Use of Formal Concept Analysis for Characterizing the Behavior of
the Residents of Bangladesh Regarding the COVID-19 Pandemic
Mateus Sobreira
a
, Martha Dias
b
, Luiz Zarate
c
and Mark Song
d
Universidade Cat
´
olica de Minas Gerais, Computer Science Department, Minas Gerais, Brazil
Keywords:
Formal Concept Analysis, COVID-19, Bangladesh, Behavior Analysis.
Abstract:
COVID-19 reported its first case on December 31, 2019, an illness rapidly spreading, becoming a global pan-
demic. Due to transmission through contact with respiratory droplets from infected individuals, governments
implemented various preventive measures to minimize the spread of the disease. This work utilized Formal
Concept Analysis (FCA) to assess the behavior of the inhabitants of Bangladesh regarding COVID-19, based
on responses to a public questionnaire conducted virtually in 64 districts from April 1 to 10, 2020. A prepro-
cessing stage was performed on the data to fit them into the format of a formal context with objects, attributes,
and the relationships among them. From the resulting general formal context, it was possible to subdivide it
according to the respondent gender, enabling an analysis of the behaviors of the men and women. Based on
these formal contexts, association rules containing the most predominant relationships in the database were
filtered using thresholds of 40% support and 80% confidence.
1 INTRODUCTION
The acute respiratory syndrome coronavirus 2
(SARS-CoV-2) reported its first case on December
31, 2019, in Wuhan, China, and spread rapidly, tak-
ing on global proportions (Chowdhury et al., 2022).
The World Health Organization designated the dis-
ease as a pandemic on January 30, 2020, and it has
since been recognized as a public health emergency
of international concern.
The transmission of the disease primarily occurs
through direct contact with an infected individual via
inhalation of respiratory droplets. Once the infection
sets in, many patients may develop an asymptomatic
condition, while others may present with mild, mod-
erate, or severe symptoms, which can potentially
progress to fatal outcomes.
Based on the advancements in understanding the
transmission mode of the disease, various countries
adopted and encouraged measures to contain the virus
to reduce the proliferation of COVID-19.
Chowdhury et al. (2022) discusses in their work
some of the measures implemented by the govern-
a
https://orcid.org/0009-0002-8983-4935
b
https://orcid.org/0000-0002-0520-9976
c
https://orcid.org/0000-0002-3018-888X
d
https://orcid.org/0000-0001-5053-5490
ment in Bangladesh, which is the focus of this study,
aimed at assisting in the prevention of COVID-19.
Campaigning for mask-wearing, adherence to social
distancing, and use of hand sanitizer for hand hygiene
were among the recommended measures.
On March 8, 2020, Bangladesh confirmed the first
case of a patient infected with SARS-CoV-2, and, ac-
cording to Ahmed and Rahman (2022), the peak of
cases and deaths resulting from the disease occurred
in June of the same year.
On October 30, 2020, Bangladesh had confirmed
404,760 positive cases of the virus, of which 5,886
individuals did not survive the syndrome.
Based on this data, the objective of this study is
to conduct a behavioral analysis of the residents of
Bangladesh regarding the COVID-19 pandemic, an-
alyzing data on preventive actions such as regular
handwashing and covering the nose and mouth when
sneezing or coughing.
Formal Concept Analysis (FCA) is used in this
study to extract inference rules that aid in understand-
ing these behaviors.
Sobreira, M., Dias, M., Zarate, L. and Song, M.
The Use of Formal Concept Analysis for Characterizing the Behavior of the Residents of Bangladesh Regarding the COVID-19 Pandemic.
DOI: 10.5220/0013139300003911
Paper published under CC license (CC BY-NC-ND 4.0)
In Proceedings of the 18th International Joint Conference on Biomedical Engineering Systems and Technologies (BIOSTEC 2025) - Volume 2: HEALTHINF, pages 475-482
ISBN: 978-989-758-731-3; ISSN: 2184-4305
Proceedings Copyright © 2025 by SCITEPRESS – Science and Technology Publications, Lda.
475

Table 1: Formal Context Example.
attribute1 attribute2 attribute3
Record 1 X
Record 2 X
Record 3 X X
Record 4 X
2 BACKGROUND
2.1 Formal Concept Analysis
The Formal Concepts Analysis (FCA) theory is an ap-
plied mathematical field whose main objective is to
represent and extract knowledge (Ganter and Wille,
1999). But it is also used for representation, informa-
tion handling, and data analysis, given the attention
received in the Data Mining field (Carpineto and Ro-
mano (2004)).
According to
ˇ
Skopljanac Ma
ˇ
cina and Bla
ˇ
skovi
´
c
(2014), this method is based on lattice theory and set
theory.
Developed by Rudolf Wille in the early 1980s, it
employs concepts interpretation as units formed over
a set of objects and a set of shared attributes. FCA
comprises three fundamental elements: formal con-
text, formal concept, and rules.
FCA is a technique based on formalizing the no-
tion of concept and structuring concepts in a concep-
tual hierarchy. The hierarchization of concepts ex-
tracted from data makes FCA a valuable tool for de-
pendency analysis.
With the increase in cases in the healthcare area
and due to the large amount of data generated,
the study and improvement of techniques to extract
knowledge are becoming increasingly justified.
FCA also permits data analysis through associa-
tions and dependencies of attributes and objects for-
mally described from a dataset.
2.1.1 Formal Context
Formally, a formal context is formed by a triple
(G, M, I), where G is a set of objects (rows), M is a
set of attributes (columns) and I is defined as the bi-
nary relationship (incidence relation) between objects
and their attributes where I ⊆ G × M.
Table 1 exemplifies a formal context. In this ex-
ample, objects correspond to tweets, attributes are the
characteristics (terms), and the relationship of inci-
dence represents whether or not the tweet has that
characteristic. An
′
X
′
is present in the table if the
tweet possesses the corresponding characteristic.
2.1.2 Formal Concepts
Let (G, M, I) be a formal context, A ⊆ G a subset of
objects, and B ⊆ M a subset of attributes. Formal
concepts are defined by a pair (A, B) where A ⊆ G is
called extension and B ⊆ M is called intention.
This pair must follow the conditions where A = B
′
and B = A
′
(Ganter and Wille, 1999). The relation is
defined by the derivation operator (
′
):
A
′
= { m ∈ M| ∀ g ∈ A, (g, m) ∈ I}
B
′
= { g ∈ G| ∀ m ∈ B, (g, m) ∈ I}
If A ⊆ G, then A
′
is a set of attributes common to
the objects of A. The derivation operator (
′
) can be
reapplied in A
′
resulting in a set of objects again (A
′′
).
Intuitively, A
′′
returns the set of all objects that
have in common the attributes of A
′
; note that A ⊆
A
′′
. The operator is similarly defined for the attribute
set. If B ⊆ M, then B
′
returns the set of objects that
have the attributes of B in common. Thus, B
′′
returns
the set of attributes common to all objects that have
the attributes of B in common; consequently, B ⊆ B
′′
.
As an example, using Table 1, ob-
jects A = {record2, record3} result in A
′
= {attribute2, attribute4} when submit-
ted to the operator described above. So
({record2, record3}, {attribute2, attribute4}) is
a concept. All concepts found from Table 1 are
displayed in Table 2.
In Table 2 there is a concept with an empty at-
tribute set and a concept with an empty object set.
They are called infimum and supremum, respectively.
2.1.3 Association Rules and Implications
The objective of this study in utilizing Formal Con-
cept Analysis lies in the extraction of rules that de-
scribe the relationships between the attributes of ob-
jects within the formal context. The association rules,
denoted as A −→ B, indicate that the objects contain-
ing the attributes in set A also include the attributes in
set B.
From the inference of the association rules, it is
possible to calculate their support, which represents
the percentage of coverage of this rule relative to the
total set of objects in the formal context. Another met-
ric derived from the association rules is their confi-
dence, which indicates the percentage of objects that
contain the attributes in set A and also possess the at-
tributes in set B.
Support and confidence are given by Equations 1
and 2, respectively.
support (r) =
|
A
′
∩ B
′
|
|
G
′
|
(1)
HEALTHINF 2025 - 18th International Conference on Health Informatics
476

Table 2: Existing concepts in the formal context of Table 1.
Objects Attributes
{Record 1, Record 2, Record 3, Record 4} {}
{Record 4} {Attribute 3}
{Record 1, Record 3} {Attribute 1}
{Record 2, Record 3} {Attribute 2, Attribute 4}
{} {Attribute 1, Attribute 2, Attribute 3, Attribute 4}
con f idence (r) =
|
A
′
∩ B
′
|
|
A
′
|
(2)
When an association rule has a confidence indica-
tor of 100%, it is said to be an implication rule.
Those inference rules extracted by Formal Con-
cept Analysis are used only to describe the character-
istics of said formal context, therefore it is not correct
to generalize their aspects to other contexts, even if
similar.
2.2 Lattice Miner
In seeking the analysis of the formal context and ex-
tracting their inference rules, this study utilized the
Lattice Miner data mining tool, version 2.0, released
in April 2017. This open-source tool is based on
the Java platform and was developed by the LARIM
research laboratory at the Universit
´
e du Qu
´
ebec en
Outaouais under the supervision of Professor Rokia
Missaoui.
One of the primary features of the Lattice Miner
tool is its ability to deliver a set of low-level opera-
tions that allow for the manipulation of input data and
association rules. This tool facilitates the generation
of groups, commonly referred to as formal concepts,
and the inclusion of their logical implications, thereby
visually representing the binary relationship between
the collection of objects and the set of attributes or
properties. The software offers the following features:
1. Refinement and Approximation
2. Generation of Formal Concepts
3. Association Rule Extraction
4. Data Import and Preparation
5. Visualization and Exploration
3 RELATED WORK
As mentioned, researchers use Formal Concept Anal-
ysis as a tool for knowledge representation, infor-
mation handling, and data analysis. Below, we de-
Table 3: 5-point Likert scale.
Value Description
1 Strongly agree
2 Agree
3 Neither agree or disagree
4 Disagree
5 Strongly disagree
scribe studies that use FCA for extracting rules from
datasets.
In the work of Salahuddin et al. (2018), a method
for image sampling is proposed to improve the results
of machine learning algorithms concerning breast
cancer.
This new method, termed Hyper Conceptual Sam-
pling (HCS), utilizes concepts from FCA to ensure
that the image retains all original information during
data filtering. Besides, no prior knowledge of the dis-
tribution of values in the original dataset is required.
A comparison using cross-validation to validate
the HCS compared its results against the Spectral-
spatial processing (SSP), proposed by Liang et al.
(2017).
The results demonstrated that the HCS method
achieved better accuracy and F1-score than the SSP
method. In one of the results, the accuracy of HCS
surpassed that of SSP by 2% when employing the
SVM technique.
Meanwhile, in the work of Jay et al. (2013),
FCA is proposed to group care trajectories based
on the sequence of patient hospitalizations in France
using data from a nationwide information system
named Programme de M
´
edicalisation des Syst
`
emes
d’Information (PMSI).
From the formal concepts, it was possible to ex-
tract an analysis focusing on two types of breast dis-
eases and the methods used to combat them, resulting
in insights regarding average expenditures, mortality
rates, and average lengths of hospitalization.
With the data from this analysis, it was possi-
ble to observe that individuals with invasive neo-
plasms receiving palliative care had the highest val-
ues across all indicators, with an average expenditure
of C26,139, a mortality rate of 69%, and an average
length of hospitalization of up to 43 days.
The Use of Formal Concept Analysis for Characterizing the Behavior of the Residents of Bangladesh Regarding the COVID-19 Pandemic
477

The authors concluded that this analysis could as-
sist healthcare professionals and the government in
resource allocation planning and efforts based on pa-
tient data rather than relying solely on visits.
Meanwhile, the work of Miranda et al. (2024)
presents an analysis of the sociocultural factors and
their contribution to the behavior of Chinese drivers.
They used FCA to extract inference rules that artic-
ulate the most relevant relationships between the ob-
jects and their attributes within the dataset.
The dataset in question consists of responses from
drivers to a self-questionnaire regarding traffic prac-
tices. The rules extracted from the used dataset identi-
fied characteristics of aggressive driving behavior re-
lated to external factors, such as friends and family,
based on the ”Mind-sponge theory”.
The study concluded that Formal Concept Anal-
ysis can help characterize these issues and assist in
decision-making to understand the root of the prob-
lem of dangerous driving.
In summary, the literature shows how FCA can
be applied in many different contexts (Ananias et al.
(2021), Alves et al. (2023)).
4 METHODOLOGY
4.1 Database
The dataset used for this study, presented in Pakpour
et al. (2020), encompasses knowledge, prevention
measures, psychological consequences, and suicidal
tendencies related to the COVID-19 pandemic in
Bangladesh.
The values presented in the dataset are responses
from a publicly available questionnaire conducted vir-
tually in 64 districts with the assistance of the Depart-
ment of Public Health and Informatics of Bangladesh,
and it was collected from April 1 to April 10, 2020.
The dataset is partitioned into sets of questions or-
ganized into eight main themes:
• Sociodemographic Information.
• Knowledge about COVID-19.
• Behaviors related to COVID-19.
• Quarantine and Economic Issues.
• Fear-related Questions about COVID-19.
• Insomnia-related Questions among Participants.
• Depression-related Questions among Participants.
• Suicidal Thoughts related to COVID-19.
In total, 11,000 people responded to the question-
naire, of which 10,067 individuals qualified for meet-
ing the three essential requirements:
• Be Bengali.
• Reside in Bangladesh.
• Be over ten years of age.
4.2 Preprocessing Step
For performing the Formal Concept Analysis, it is
necessary to preprocess the database to allow the for-
mal context construction.
In the initial step of data preprocessing, we con-
ducted a thorough analysis and explored the available
attributes to identify those irrelevant to the subject of
the work or that contain empty values. The primary
purpose of this stage is to reduce the dimensionality of
the database to facilitate the use of the FCA approach.
To obtain the implication rules, the core ele-
ment of this work, it is crucial to binarize and dis-
cretize the previously selected attributes. Some of
these attributes have categorical values representing
the agreement level expressed by respondents on cer-
tain questions or the likelihood of performing specific
actions.
The categorical values have their ranges measured
by the Likert Scale (Sullivan and Artino Jr, 2013).
The original dataset has categorical questions valued
by the 5-point Likert scale, where respondents show
their agreement or disagreement level with the state-
ments. The values from this range are ordinal.
Table 3 shows the range from the 5-point Likert
scale.
4.2.1 COVID-19 Behavior Questions
The set of attributes related to the COVID-19 Behav-
ior Questions (BRQ) requires the binarization process
of its elements for applying the FCA method. These
attributes present five possible answers to describe the
likelihood of the respondent engaging in the behavior
in question as in the 5-point Likert Scale. The possi-
ble responses are:
• Never.
• Seldom.
• Sometimes.
• Often.
• Almost always.
The binarization process for this set of attributes
entails transforming the responses given as less likely
end (never and rarely) into a negative result, while
HEALTHINF 2025 - 18th International Conference on Health Informatics
478

the remaining (sometimes, usually, and always) into
a positive answer. Figure 1 shows this characteriza-
tion.
Figure 1: Preprocess steps on Behaviour Related Questions.
4.2.2 Fear of COVID-19 Scale
Another set of attributes in the database that require
the binarization process are the questions related to
the Fear of COVID-19 Scale (FCV-19S), which, simi-
larly to the previously described set, also presents five
possible responses to the questions.
However, these pertain to the agreement or dis-
agreement level of the respondent regarding certain
situations that could occur during the COVID-19 pan-
demic. The possible responses to these questions are:
• Strongly agree.
• Agree.
• Neither agree or disagree.
• Disagree.
• Strongly disagree.
In this set of attributes, the binarization process
initially focused on removing the responses that con-
tained a neutral stance on the questions posed (nei-
ther agree nor disagree). This removal facilitates the
separation of the responses into two groups, where
the group with a higher degree of disagreement (dis-
agree and strongly disagree) was transformed into a
negative result, while the group with a higher degree
of agreement (agree and strongly agree) transformed
into a positive one. Figure 2 presents this process.
4.3 Formal Context
From the stages of preprocessing the database, we
obtained three formal contexts whose we study their
inference rules through Formal Concept Analysis.
These contexts are from the same set of 12 attributes,
in which:
• 2 attributes refer to sociocultural aspects.
• 3 attributes refer to knowledge about COVID-19.
Figure 2: Preprocess steps on Fear of COVID-19 Scale.
• 4 attributes refer to behaviors adopted against
COVID-19.
• 3 attributes refer to fears related to COVID-19.
The general formal context comprises all the re-
sponses obtained through the preprocessing stages,
totaling 6,067 objects. The other two formal con-
texts are subsets of the general formal context, with
the male formal context consisting of all objects iden-
tifying as male, totaling 3,351 objects.
Figure 3: Preprocess steps to create the Formal Contexts.
The Use of Formal Concept Analysis for Characterizing the Behavior of the Residents of Bangladesh Regarding the COVID-19 Pandemic
479

Table 4: Part of the extracted Formal Context.
Single Married I’m afraid of COVID-19 I stay in my house when not feeling well
Person 1 × × ×
Person 2 × × ×
Person 3 × ×
Person 4 × × ×
Person 5 × × ×
Table 5: Rules extracted in the Broad Formal Context.
Rule Support Confidence
IF I believe that COVID-19 spreads through coughs and sneezes,
I believe that the mask use should be mandatory,
when sneezing I cover mouth and nose,
thinking about COVID-19 makes me uncomfortable,
I stay in my house if I’m not feeling well and
I maintain a 1 meter distance between people that are coughing or sneezing
THEN I’ll wash my hands periodically.
50% 99%
IF I believe that the mask use should be mandatory,
when sneezing I cover mouth and nose,
thinking about COVID-19 makes me uncomfortable,
I wash my hands periodically,
I am afraid of COVID-19 and
I’m afraid of losing my life for COVID-19
THEN I’ll stay in my house when I’m not feeling well.
40% 96%
IF I believe that the mask use should be mandatory,
thinking about COVID-19 makes me uncomfortable,
I wash my hands periodically,
I maintain a 1 meter distance between people that are coughing or sneezing,
I stay in my house when I’m not feeling well and
I’m afraid of COVID-19
THEN I’ll cover my nose and mouth when sneezing.
51% 98%
The female formal context comprises all objects
identifying as female, totaling 2,716. Figure 3 illus-
trates the process for constructing the formal context,
while Table 4 presents a portion of the overall formal
concept.
5 RESULTS AND DISCUSSION
To analyze the most significant inference rules from
the three formal contexts, we apply threshold values
for support and confidence to recognize these rules.
In this work, we evaluate only rules with values
exceeding 40% of support and 80% confidence.
Lattice Miner stored these rules into an XML file.
We analyzed them in the form A −→ B, where A rep-
resents the antecedent and B represents the conse-
quent.
In the overall scenario, represented by the formal
context containing all eligible objects after the prepro-
cessing stage of the original database, a total of 2,828
inference rules were extracted. Table 5 shows some
of these rules.
From the rules presented in Table 5, it is evi-
dent that individuals who already engaged in actions
against COVID-19, such as covering their nose and
mouth when sneezing, maintaining social distancing
of at least 1 meter from someone who is coughing or
sneezing, and who feel uncomfortable or fearful about
COVID-19, will regularly wash their hands. This in-
ference rule exhibited 50% support and 99% confi-
dence.
Another rule that can be analyzed from the data in
Table 5 states that individuals who engaged in actions
against COVID-19, such as regularly washing their
hands and covering their nose and mouth when they
are sneezing and who feel uncomfortable and afraid
HEALTHINF 2025 - 18th International Conference on Health Informatics
480

Table 6: Rules extracted in the Male Formal Context.
Rule Support Confidence
IF I’m single and
I’m afraid of COVID-19
THEN I’ll wash my hands periodically.
52% 98%
IF I’m single,
I believe that COVID-19 spreads out by coughs or sneezes,
I believe that masks should be mandatory,
I wash my hands periodically and
I stay in my house when I’m not feeling well
THEN I’ll cover my mouth and nose when sneezing.
56% 95%
Table 7: Rules extracted in the Female Formal Context.
Rule Support Confidence
IF I’m single and
I’m afraid of COVID-19
THEN I’ll wash my hands periodically.
54% 98%
IF I’m single,
I believe that COVID-19 spreads out by coughs or sneezes,
I believe that masks should be mandatory,
I wash my hands periodically and
I stay in my house when I’m not feeling well
THEN I’ll cover my mouth and nose when sneezing.
52% 96%
of dying from COVID-19, will stay at home in the
event of any illness. This inference rule demonstrated
40% support and 96% confidence.
Regarding the scenario that focuses exclusively on
objects with the attribute of male gender, a total of
1,722 inference rules were extracted. In contrast, the
scenario with only objects identifying as female gen-
der yielded 4,054 rules. Tables 6 and 7 present some
of these rules, representing objects with male and fe-
male genders, respectively.
A discrepancy in the number of inference rules ex-
tracted by the Lattice Miner is evident in the two sce-
narios described above.
This difference is attributed to the variation in the
number of attributes in the two formal contexts since
fewer objects with related attributes are needed for the
rule to exceed the acceptance thresholds of support
and confidence.
Upon analyzing the inference rules present in Ta-
bles 6 and 7, it is possible to observe that single males
who believe that COVID-19 spreads through cough-
ing or sneezing and engage in actions against COVID-
19, such as regularly washing their hands, will cover
their nose and mouth less when sneezing than fe-
males.
Support and confidence indices from these rules
show this difference by comparison. For men, these
indices correspond to 56% support and 95% confi-
dence. Meanwhile, for women, 52% support and 96%
confidence.
6 CONCLUSION
In this work, we applied Formal Concept Analysis to a
database from Bangladesh. This dataset has attributes
regarding the knowledge, preventive measures, and
psychological consequences among the inhabitants of
this country.
So, FCA was able to extract inference rules
for studying the population’s behavior regarding the
COVID-19 pandemic. With this data, it was also pos-
sible to conduct a more detailed comparison of the
behaviors of females and males.
From the most relevant extracted inference rules,
it was possible to relate actions and psychological as-
pects concerning the COVID-19 pandemic with be-
haviors adopted by the population of Bangladesh,
such as regular handwashing and the adoption of self-
imposed quarantine in the event of the onset of any
illness. These rules were obtained using a minimum
acceptance threshold of 40% support and 80% confi-
dence.
It is possible to identify two issues within the
The Use of Formal Concept Analysis for Characterizing the Behavior of the Residents of Bangladesh Regarding the COVID-19 Pandemic
481

database used: one concerning its geographical lim-
itation and the other regarding the imprecision of the
data source.
The issue of geographical limitation arises from
the fact that the data represented pertains solely to the
inhabitants of Bangladesh, which restricts the gener-
alizability of the rules found in this work.
The problem of data imprecision is related to the
difficulty of characterizing more specific and individ-
ual aspects of participants due to the subjectivity and
bias of the responses, as they are part of a public self-
assessment questionnaire.
For future studies, I propose conducting an up-
dated version of the public research presented in
Pakpour et al. (2020), so that it will be possible to
extract its inference rules and compare the most im-
portant relationships found, aiming to highlight the
behavioral changes of the inhabitants of Bangladesh
during the COVID-19 pandemic and in the post-
pandemic period.
It would also be relevant in future works, to im-
plement other data extraction methods, such as clus-
ter analysis, for a comparative evaluation of their re-
sults with the inference rules obtained through Formal
Concept Analysis.
ACKNOWLEDGEMENTS
The authors thank the Pontif
´
ıcia Universidade
Cat
´
olica de Minas Gerais – PUC-Minas and
Coordenac¸
˜
ao de Aperfeic¸oamento de Pessoal de
N
´
ıvel Superior — CAPES (CAPES – Grant PROAP
88887.842889/2023-00 – PUC/MG, Grant PDPG
88887.708960/2022-00 – PUC/MG - Inform
´
atica,
and Finance Code 001). The present work was also
carried out with the support of Fundac¸
˜
ao de Amparo
`
a Pesquisa do Estado de Minas Gerais (FAPEMIG)
under grant number APQ-01929-22.
REFERENCES
Ahmed, A. and Rahman, M. M. (2022). Critical analysis of
the possible determining factors of covid-19 epidemi-
ological trend in bangladesh: A review. Bangladesh
Journal of Infectious Diseases, 8(1):36–41.
Alves, A., Zarate, L., Freitas, H., and Song, M. (2023). Par-
allelism in the generation of concepts through the for-
mal context object partitioning using the in-close 4 al-
gorithm. pages 195–202.
Ananias, K. H., Missaoui, R., Ruas, P. H., Zarate, L. E., and
Song, M. A. (2021). Triadic concept approximation.
Information Sciences, 572:126–146.
Carpineto, C. and Romano, G. (2004). Concept Data Anal-
ysis: Theory and Applications. John Wiley & Sons
Inc.
Chowdhury, N. M., Farrque, M. M. O., Hassin, C. N., and
Henderson, A. (2022). Community survey on wearing
mask to prevent covid-19, bangladesh, 2021. Inter-
national Journal of TROPICAL DISEASE & Health,
43(12):9–15.
Ganter, B. and Wille, R. (1999). Formal concept analysis :
mathematical foundations. Springer.
Jay, N., Nuemi, G., Gadreau, M., and Quantin, C. (2013).
A data mining approach for grouping and analyzing
trajectories of care using claim data: the example of
breast cancer. BMC Medical Informatics and decision
making, 13:1–9.
Liang, J., Zhou, J., Qian, Y., Wen, L., Bai, X., and Gao,
Y. (2017). On the sampling strategy for evaluation of
spectral-spatial methods in hyperspectral image clas-
sification. IEEE, 55(2):862–880.
Miranda, D., Z
´
arate, L. E., and Song, M. A. (2024). Ap-
plication of formal concept analysis to characterize
driving behaviors and socio-cultural factors related to
driving. In BIOSTEC (2), pages 56–62.
Pakpour, A. H., Al Mamun, F., Hosen, I., Griffiths,
M. D., and Mamun, M. A. (2020). A population-
based nationwide dataset concerning the covid-19
pandemic and serious psychological consequences in
bangladesh. Data in Brief, 33:106621.
Salahuddin, T., Haouari, F., Islam, F., Ali, R., Al-Rasbi, S.,
Aboueata, N., Rezk, E., and Jaoua, A. (2018). Breast
cancer image classification using pattern-based hyper
conceptual sampling method. Informatics in Medicine
Unlocked, 13:176–185.
Sullivan, G. M. and Artino Jr, A. R. (2013). Analyzing and
interpreting data from likert-type scales. Journal of
graduate medical education, 5(4):541–542.
ˇ
Skopljanac Ma
ˇ
cina, F. and Bla
ˇ
skovi
´
c, B. (2014). Formal
concept analysis – overview and applications. Proce-
dia Engineering, 69:1258–1267. 24th DAAAM Inter-
national Symposium on Intelligent Manufacturing and
Automation, 2013.
HEALTHINF 2025 - 18th International Conference on Health Informatics
482