Exploring Links Between Social Media Habits, Loneliness, and Sleep: A
Formal Concept Analysis Approach
Fernanda M. Gomes
a
, Julio C. V. Neves
b
, Luis Enrique Z
´
arate G
´
alvez
c
and Mark Alan Junho Song
d
Instituto de Ci
ˆ
encias Exatas e Inform
´
atica, Pontif
´
ıcia Universidade Cat
´
olica de Minas Gerais, Brazil
Keywords:
Formal Concept Analysis, Mental Health, Social Media, Sleep Quality, Data Science, Lattice Miner,
Loneliness.
Abstract:
Social media platforms have reshaped personal interactions, allowing engagement with diverse audiences.
However, growing evidence suggests that these platforms may also contribute to mental health challenges.
This paper investigates the associations between social media usage patterns, loneliness, and sleep quality,
using Formal Concept Analysis (FCA) on data from a sample in Bangladesh. The dataset includes information
on social media habits, loneliness, anxiety, depression, and sleep disturbances, using metrics from validated
psychological scales. Through FCA, this study extracted implication rules that describe how specific social
media usage behaviors relate to feelings of loneliness and sleep issues. Findings show that individuals with
high levels of social media engagement report shorter sleep durations and heightened symptoms of loneliness.
FCA is used in this study to uncover non-obvious relationships within complex datasets, making it a valuable
approach for analyzing patterns between social media behaviors and mental health outcomes.
1 INTRODUCTION
Social media are online platforms in which individu-
als can engage with others and present themselves in a
chosen manner. These platforms allow for interaction
with a range of audiences, from small groups to vast
networks, providing value through content created by
users and the feeling of social connection (Carr and
Hayes, 2015).
This research aims to uncover links between lone-
liness, social media usage, and sleep patterns by ap-
plying Formal Concept Analysis (FCA) to a publicly
available dataset from Bangladesh. The data pro-
vided include information on Social Networking Sites
(SNS) usage and validated mental health scales. FCA
was introduced in 1982 by Rudolf Wille as a deriva-
tion of concept hierarchy from a set of objects and
their properties (Wille, 2009). FCA has received sig-
nificant interest in fields such as health, software en-
gineering, and data mining.
Previous studies, such as Marttila’s (Marttila et al.,
a
https://orcid.org/0009-0005-5607-1689
b
https://orcid.org/0000-0002-0520-9976
c
https://orcid.org/0000-0001-7063-1658
d
https://orcid.org/0000-0001-7315-3874
2021), suggest that “increased problematic social me-
dia use (PSMU) predicts increased individuals lone-
liness over time, as well as that increased loneliness
predicts decreased life satisfaction”. It is interesting
to note the hypothesis that “in some cases, PSMU
might initiate a cycle of social comparison, isolation,
and decreased life quality, happiness, and life satis-
faction”.
Furthermore, Alonzo (Alonzo et al., 2021) notes
that “longitudinal studies suggest poor sleep quality
and frequent sleep disturbances may partially explain
the association between excessive social media use
and poor mental health outcomes.”
Examining these relationships through Formal
Concept Analysis enable us to move beyond sim-
ple correlations and to explore the interplay between
social media behaviors and mental health outcomes.
Understanding these dynamics is essential for pub-
lic health interventions, developing responsible plat-
form design, and empowering individuals to cultivate
healthier relationships with social media.
This paper is structured as follows: Section 2 in-
troduces the background of the study; Section 3 is
about related work; Section 4 describes the methodol-
ogy; Section 5 presents and discusses the results and
Section 6 outlines the conclusion and future work.
226
Gomes, F. M., Neves, J. C. V., Gálvez, L. E. Z. and Song, M. A. J.
Exploring Links Between Social Media Habits, Loneliness, and Sleep: A Formal Concept Analysis Approach.
DOI: 10.5220/0013267600003929
Paper published under CC license (CC BY-NC-ND 4.0)
In Proceedings of the 27th International Conference on Enterprise Information Systems (ICEIS 2025) - Volume 1, pages 226-233
ISBN: 978-989-758-749-8; ISSN: 2184-4992
Proceedings Copyright © 2025 by SCITEPRESS – Science and Technology Publications, Lda.

2 BACKGROUND
2.1 Social Network - Mental Health
Over the past decade, the extensive use of Social Net-
working Sites (SNS) like Instagram, Facebook, and
Twitter has fundamentally transformed how we in-
teract, communicate, and process information. The
broad adoption of these platforms in our everyday
lives has influenced multiple aspects of modern soci-
ety, including advertising, education, public relations
and political campaigning As a result, social media
has become a fundamental element of contemporary
society.
The World Health Organization highlights psy-
chological health issues as a major contributor to de-
clining global mental health. Rising rates and severity
of mental illnesses emphasize the need for prioritiz-
ing mental health planning, despite often being over-
looked. Mental illnesses are among the top causes of
disability for individuals aged 15 to 44, with many
other disabilities also linked to mental health chal-
lenges (Kumar et al., 2024).
Among mental health concerns, loneliness has
emerged as a pressing issue in the digital age, with
its implications far-reaching across demographics and
social contexts. The widespread adoption of SNS
has reshaped social interactions, enabling connections
but sometimes heightening feelings of isolation. Al-
though these platforms are intended to enhance con-
nectivity, their impact on users’ mental health is com-
plex.
2.2 Formal Concept Analysis
Formal Concept Analysis (FCA) is a method used
to identify patterns by employing association rules
and their implications (Ganter and Wille, 2012). It
operates on the principle that concepts are formed
through the relationships between objects and at-
tributes, which allows for the development of concep-
tual hierarchies and facilitates a deeper understanding
of associations between significant terms. As such,
FCA proves especially valuable for organizing and
extracting insights from data sets.
This technique is often applied in studies that ex-
amine domains represented as binary tables of objects
and attributes. While longitudinal studies investigate
a group of individuals with specific traits over mul-
tiple time periods (referred to as waves), FCA dif-
fers in that it concentrates on analyzing the seman-
tic structure of data at a single point in time, without
accounting for temporal changes. Therefore, FCA is
particularly useful when the goal is to understand the
underlying structure of data at a given moment.
At the core of FCA is the concept of a formal con-
text, represented as a triple K = (G, M, I), where G is
a set of objects, M is a set of attributes, and I ⊆ G × M
is the incidence relation, meaning that (g, m) ∈ I indi-
cates that object g has attribute m.
Table 1: Formal Context Example.
Attribute 1 Attribute 2 Attribute 3 Attribute 4
object 1 X
object 2 X X
object 3 X X X
object 4 X
Table 1 exemplifies a formal context. In this
example, objects correspond to lines, attributes to
columns, and the relationship of incidence represents
whether or not the object has an specific characteris-
tic. An
′
X
′
is present in the table if the object pos-
sesses the corresponding characteristic.
For a set of objects A ⊆ G, the set of common
attributes shared by the objects in A is denoted as
A
′
=
{
m ∈ M | ∀g ∈ A : (g, m) ∈ I
}
. Similarly, for a
set of attributes B ⊆ M, the set of objects sharing these
attributes is B
′
=
{
g ∈ G | ∀m ∈ B : (g, m) ∈ I
}
.
Building on this foundation, a formal concept in
a formal context K = (G, M, I) is defined as a pair
(A, B), where A is called the extension (the set of ob-
jects) and B the intention (the set of attributes). For a
pair (A, B) to qualify as a concept, it must satisfy the
conditions A = B
′
and B = A
′
. The set of all formal
concepts in context K is denoted as β(K).
As an example, using Table 1, objects
A = {ob ject2, ob ject3}, when submitted to
the operator (
′
) described above, will re-
sult in A
′
= {attribute2, attribute4}. So
{{ob ject2, ob ject3}, {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.
Furthermore, FCA supports the generation of as-
sociation rules that help highlight relationships within
the data. Rules are dependencies between elements of
a set obtained from a formal context.
Given the context (G, M, I) the rules of implica-
tion are of the form B → C if and only if B, C ⊂ M
and B
′
⊂ C
′
. A rule B → C is considered valid if and
only if every object that has the attributes of B will
also have the attributes of C.
Given a rule r and parameters s and c, one can
denote:
s = suppr (r) =
|
A
′
∩ B
′
|
|
G
′
|
(1)
Exploring Links Between Social Media Habits, Loneliness, and Sleep: A Formal Concept Analysis Approach
227

Table 2: Existing concepts in the formal context of Table 1.
Objects Attributes
{object 1, object 2, object 3, object 4} {}
{object 4} {attribute 3}
{object 1, object 3} {attribute 1 }
{object 2, object 3} {attribute 2, attribute 4}
{} {attribute 1, attribute 2, attribute 3, attribute 4}
- known as the support of the rule r, and
c = con f (r) =
|
A
′
∩ B
′
|
|
A
′
|
(2)
- referred to as confidence.
These are key metrics for evaluating association
rules. Support (s) represents the proportion of trans-
actions in which both attributes A ∪ B appear, relative
to the total number of transactions. It indicates how
often a rule is applicable within the dataset. Confi-
dence (c), on the other hand, measures the likelihood
that if a transaction contains A (the antecedent), it will
also include B (the consequent), expressed as a per-
centage or fraction.
3 RELATED WORK
Formal Concept Analysis (FCA) has been recognized
for its ability to handle complex data and extract
meaningful relationships.
In (
ˇ
Skopljanac Ma
ˇ
cina and Bla
ˇ
skovi
´
c, 2014) the
authors provide an overview of FCA’s theoretical
foundations and its applications across various fields,
including computer-aided learning, information re-
trieval and machine learning. Their work shows
how FCA can uncover patterns and hierarchies within
large datasets, making it a valuable tool for knowl-
edge representation and analysis in different domains,
including education and e-learning systems.
FCA has also proven to be a versatile tool across
multiple fields of study (Poelmans et al., 2013). A
comprehensive survey of FCA applications was con-
ducted, highlighting its usage in areas like text, web
and software mining, life sciences and ontology engi-
neering.
(Lana et al., 2022) perform a longitudinal analysis
of a COVID-19 database using triadic formal concept
analysis. The study presents implication rules that
describe the evolution of the COVID-19 pandemic
across different points in time.
A longitudinal study provides valuable insights
into the evolution of psychological behaviors, espe-
cially during the COVID-19 pandemic. (Coutinho
et al., 2024) applied triadic analysis, based on For-
mal Concept Analysis (FCA), to examine a longitu-
dinal dataset capturing individuals’ attitudes and re-
actions throughout the pandemic. By deriving rules,
the work illustrates how various factors interact under
different pandemic conditions, revealing stress lev-
els associated with disease prevention measures. The
findings emphasize how these stress-related behaviors
evolved, offering a nuanced view of psychological re-
sponses across different pandemic scenarios.
(Song et al., 2024) applied triadic formal concept
analysis (FCA) to characterize infant mortality across
different regions of Minas Gerais, Brazil. Key factors
identified included birth weight, gestation period, and
APGAR scores. The findings revealed associations
among these variables, underscoring the significance
of maternal education and prenatal care consultations.
Furthermore, FCA has been recognized as an ef-
fective tool for analyzing complex, unstructured data,
and has demonstrated its value in both theoretical and
practical applications, such as gene expression analy-
sis and evaluating chemical compound properties.
By leveraging FCA, the present study focuses
on extracting meaningful patterns and associations
within the selected dataset, contributing to the on-
going exploration of FCA’s applicability in health-
related domains.
While traditional FCA is based on a dyadic struc-
ture, consisting of objects and attributes, Triadic
Concept Analysis (TCA) introduces a third element-
conditions, forming triadic contexts (Wille, 1995).
These triadic structures are mathematically repre-
sented as complete trilattices, allowing for more com-
plex relationships between the three components to be
explored. TCA enables the comprehension of sets of
concepts within a three-dimensional context, offering
a framework for data analysis in fuzzy environments
(Lehmann and Wille, 1995).
4 METHODOLOGY
This section details the methodology, including the
tool employed for Formal Concept Analysys, the
ICEIS 2025 - 27th International Conference on Enterprise Information Systems
228

dataset selected and the data preprocessing steps ap-
plied. The complete process is represented in Figure
2.
4.1 Database
The database used in this paper is the Data set con-
cerning the use of social networking sites and men-
tal health problems among the young generation in
Bangladesh, publicly available on Science Direct (Is-
lam et al., 2021). The study was conducted by
researchers from the University of Asia Pacific in
Dhaka between February and March 2021, using
Google Forms.
Initially, the survey received 826 responses, how-
ever, 35 were excluded due to partial or incomplete
information. This resulted in a final sample of 791
adults from Bangladesh, aged between 15 and 40
years. The study included participants from differ-
ent education levels, economic statuses, and occupa-
tions, ensuring that the analysis captures varied per-
spectives.
The survey aimed to explore the relationship be-
tween social networking site (SNS) use and four di-
mensions of psychological distress: depression, anxi-
ety, loneliness, and sleep disturbances.
It was divided into sections focusing on sociode-
mographic information, SNS usage patterns and as-
sessments of mental health problems using interna-
tionally validated scales - UCLA Loneliness Scale,
PHQ-9 (depression), GAD-7 (anxiety) and PSQI
(sleep quality).
The UCLA Loneliness Scale is a tool assess sub-
jective feelings of loneliness and social isolation. The
scale consists of 20 items rated on a 4-point scale,
ranging from “never” to “often” (Russell, 1996).
The PHQ-9 is a self-administered tool used to
screen for depression and measure its severity. It con-
sists of 9 questions that align with the diagnostic cri-
teria for major depressive disorder in the DSM-IV.
Each item is scored from 0 (not at all) to 3 (nearly
every day), with a total score ranging from 0 to 27.
Higher scores indicate more severe levels of depres-
sive symptoms.
Moreover,the GAD-7 is a brief self-report ques-
tionnaire used to identify generalized anxiety disor-
der (GAD). The survey includes 7 items, each scored
from 0 (not at all) to 3 (nearly every day), with a total
score ranging from 0 to 21. Scores of 10 or higher
suggest the presence of GAD.
Finally, the Pittsburgh Sleep Quality Index (PSQI)
is a self-report questionnaire that measures sleep qual-
ity over a one-month period. It has 19 items that gen-
erate seven component scores: subjective sleep qual-
ity, sleep latency, sleep duration, habitual sleep effi-
ciency, sleep disturbances, use of sleep medication,
and daytime dysfunction. They are then combined to
create a global score that reflects overall sleep quality,
with higher scores indicating worse sleep outcomes
(Buysse et al., 1989).
4.1.1 Data Preprocessing
Originally, the dataset had 51 attributes spanning de-
mographics, social media usage habits and health-
related information.
For this study, sixteen attributes were selected ac-
cording to their direct relevance to studying the rela-
tionships between social media usage, loneliness, and
sleep patterns . Of these, five were created based on
the original survey by turning specific responses into
binary variables.
Table 3 provides a full list of the selected at-
tributes, while the newly created binary variables and
their coding criteria are detailed below:
• A: Main purpose of social media usage is social
communication.
For the question “What is your main purpose for
using social media?”, the response indicating “To
stay connected with people” was coded as 1 and
all other purposes were coded as 0.
• B: Uses social media at anytime of day.
For the question “When do you usually use social
media?”, the response “Frequently at anytime”
was coded as 1 and the other responses as 0.
• C: Sleeps for less than 7 hours.
For the question “How many hours of actual sleep
do you get at night?”, responses 4-6 hours and
“Less than 4 hours” were coded as 1, indicat-
ing less than the recommended 7 hours of sleep.
Other responses as 0.
• F: More than 3 hours of social media per day.
For the question “How much time do you spend
daily in social media?”, the responses were cate-
gorized in two groups: “More than 5 hours” and
“3-5 hours” were coded as 1, the rest as 0.
• H: Trying to reduce or stop the use of Social Me-
dia.
For the question “Are you trying to control that
thing and trying to reduce the use of social me-
dia?”, responses “Trying to stop the use” and
“Trying to reduce the use” were coded as 1, others
as 0.
To further prepare the dataset for Formal Concept
Analysis (FCA), binarization rules were applied to all
Exploring Links Between Social Media Habits, Loneliness, and Sleep: A Formal Concept Analysis Approach
229

Figure 2: Methodology Diagram.
Table 3: Attributes Selected.
Attribute Caption
A Main purpose of social media usage is social communication.
B Uses social media at any time of day.
C Sleeps for less than 7 hours.
D Has experienced peer pressure due to social media.
E Feels like people are around them but not with them in the last 30 days.
F Uses social media for more than 3 hours per day.
G Feels isolated from others in the last 30 days.
H Trying to reduce or stop the use of social media.
I Thinks mental well-being would improve without social media.
J Trouble concentrating on things in the last 30 days.
K Feels left out in the past 30 days.
L Feels like people are around them but not with them in the past 30 days.
M Feels a lack of companionship in the past 30 days.
N Feels tired or has little energy in the last 30 days.
O Worries too much about different things in the last 30 days.
P Feels like there is no one they can turn to in the last 30 days.
categorical variables. The full binarization schema is
presented in Table 4.
Table 4: Rules applied to the dataset.
Original Value Binarized Value
Never, Rarely 0
Sometimes, Often 1
Not at All, Several Days 0
Half Days, Nearly Everyday 1
No 0
Yes 1
4.2 Lattice Miner
Lattice Miner 2.0 is a public domain Java tool de-
veloped by Kevin Emamirad under the supervision
of Rokia Missaou at the Universit
´
e du Qu
´
ebec en
Outaouais.
It enables the representation and manipulation of
input data, concept lattices, and association rules, be-
ing a powerful tool for data analysis (Missaoui and
Emamirad, 2017).
In the context of this work, Lattice Miner was used
for extracting association rules between mental health
concerns, sleep patterns and SNS usage.
5 RESULTS
When applying FCA, the minimum thresholds for
support (sup) and confidence (conf) were set at 20%.
As previously discussed, support is the proportion of
transactions with both the antecedent and the conse-
quent. Confidence, on the other hand, measures the
likelihood that if a transaction includes A it will also
include B.
With these thresholds, 389 implication rules were
generated in the format A → B, in which A represents
the antecedent and B the consequent. The selected
rules will be discussed in detail within this section.
ICEIS 2025 - 27th International Conference on Enterprise Information Systems
230

5.1 Social Media Usage Patterns
Both rules presented in Table 5 highlight the potential
addictive nature of SNS. Despite the intention to cut
back or awareness of potential mental health benefits
from reduced usage, users frequently struggle to limit
their time on these platforms.
Table 5: Rules related to Social Media usage patterns.
Index Rule Support Confidence
01 H → F 27% 47%
02 I → F 29% 47%
Rule 01 suggests that even among users actively
trying to reduce or stop social media usage (H), nearly
half still spend more than three hours daily on SNS
(F).
Similarly, for Rule 02, among those who believe
their mental well-being would benefit from reduced
social media usage (I), a significant portion (47%)
continues to engage heavily, with over three hours of
daily use (F).
Studies show that the use of Social Networking
Sites can lead to addictive behaviors, characterized by
excessive use, withdrawal symptoms, and difficulty in
controlling time spent on these platforms.
Heavy usage is often motivated by a need for so-
cial connection and a fear of missing out, leading
users into a cycle of continued engagement even when
they recognize potential benefits of reducing screen
time (Kuss and Griffiths, 2017).
These results suggest a potential addictive pattern
in SNS use, where users, despite efforts to limit us-
age, continue to engage heavily, indicating possible
dependency behaviors
5.2 Social Media as a Social
Communication Tool
Rules displayed in Table 6 show that social motiva-
tions can be a factor that leads to heavy, unrestricted
and potentially detrimental SNS use.
Table 6: Rules associated to individuals that use SNS as a
social communication tool.
Index Rule Support Confidence
03 A → B 39% 69%
04 A → F 28% 50%
05 A → C 31% 56%
Rule 03 implies that 69% of individuals who pri-
marily use Social Media for Social communication
(A) tend to access these platforms at any given time
(B).
Furthermore, Rule 04 demonstrates that within
the same group, half of the users spend more than
three hours daily on SNS (F).
Finally, Rule 05 shows that of all the individuals
who primarily use social media for social communi-
cation (A), 56% report less than seven hours of sleep
(C).
5.3 Social Media Usage and Sleep
Table 7: Rules that reveal patterns between SNS use and
sleep.
Index Rule Support Confidence
06 B → C 36% 58%
07 F → C 30% 64%
08 C → J 20% 31%
As shown in Table 7 there was a significant relation-
ship between high SNS use and reduced sleep dura-
tion.
According to the National Sleep Foundation, the
minimum appropriate sleep duration for young adults
and adults is seven hours (Hirshkowitz et al., 2015).
Sleep deprivation impairs perception, concentra-
tion, vision, and reaction time. It also leads to poor
memory, rigid thinking, poor decision-making, and
emotional issues (Orzeł-Gryglewska, 2010).
For Rule 06, a support level of 36% alongside a
confidence level of 58% suggests that individuals who
access social media frequently at any time (B) are
more likely to report sleeping less than seven hours
per night (C).
Likewise, Rule 07 indicates that those who spend
over three hours daily on SNS (F) are more likely to
experience shorter sleep durations (C).
Longitudinal studies have found that frequent so-
cial media use was a risk factor for both poor mental
health and poor sleep outcomes (Alonzo et al., 2021).
As for the consequences of the lack of sleep, Rule
08, with a support level of 20% and a confidence of
31%, implies that individuals who state sleeping less
than seven (C) hours per night are also more likely to
experience difficulties concentrating on tasks, such as
reading or watching television (J).
These insights align with research suggesting
that limiting screen time could play a crucial role
in improving sleep quality and, consequently, men-
tal health outcomes (Exelmans and Van den Bulck,
2016).
The findings show a significant association be-
tween high SNS engagement and reduced sleep dura-
tion, aligning with previous research that links screen
time to impaired sleep quality.
Exploring Links Between Social Media Habits, Loneliness, and Sleep: A Formal Concept Analysis Approach
231

5.4 Social Media and Loneliness
Rules shown in Table 8 reveal a relationship between
extended Social Media usage and indicators of loneli-
ness , outlining a complex connection between these
factors.
Table 8: Rules that demonstrate a duality between SNS use
and loneliness.
Index Rule Support Confidence
09 F → L 26% 54%
10 F → M 24% 50%
11 F → K 23% 48%
12 F → G 23% 48%
13 L → F 26% 53%
14 M → F 24% 53%
15 K → F 23% 54%
16 G → F 23% 51%
Rules 09-12 indicate that people that spending
more than three hours on Social Media daily (F) is
associated with several loneliness-related symptoms:
• Feeling that people are around but not with them
(L);
• Lacking companionship (M);
• Feeling left out (K);
• Experiencing isolation from others (G).
Conversely, Rules 13-16 reveal that feelings of
loneliness (L, M, K, G) often lead individuals to spend
more time on social media (F). This duality sug-
gests that while social media may initially seem like a
remedy for loneliness, frequent use can deepen those
same feelings, creating a cycle where the pursuit of
social connection ultimately reinforces a sense of iso-
lation.
Supporting research shows that lonely individu-
als turn to social media to fill the gap left by lim-
ited in-person relationships, yet they often do not find
the support they seek online. Additionally, loneliness
is frequently associated with problematic social me-
dia usage patterns (O’Day and Heimberg, 2021; Song
et al., 2014), and this excessive engagement with SNS
has been linked to heightened feelings of loneliness
(Marttila et al., 2021).
This cyclical pattern suggests that while social
media can temporarily address feelings of loneliness,
frequent use may exacerbate isolation, creating a self-
reinforcing loop that can contribute to poor mental
health outcomes.
6 CONCLUSION AND FUTURE
WORK
The present study used FCA to further investigate the
associations between Social Media use, sleep patterns
and feelings of loneliness among a sample of young
adults in Bangladesh.
Among fifty one attributes, the sixteen more rele-
vant to the study were selected and binarized. In se-
quence, Lattice Miner, a public domain Java tool, was
used to extract the implication rules.
From 389 extracted rules, sixteen were selected
for in depth analysis, highlighting significant links
such as the association between extended social me-
dia use and disrupted sleep, and the cyclical relation-
ship between loneliness and increased social media
use. Here are the main outcomes:
• Users who attempt to reduce their time spent on
Social Media, often find it challenging to do so.
Even when individuals recognize potential bene-
fits of limiting their online time, the appeal of on-
going engagement with SNS endures;
• Those who primarily use SNS as a means of so-
cial communication tend to spend significant time
online, often engaging without restriction, which
can lead to reduced sleep duration;
• Frequent social media users, especially those who
access platforms at all hours, are more likely to
experience shorter sleep time, which can con-
tribute to cognitive challenges, such as difficulties
with concentration.
• High SNS engagement correlates with feelings of
isolation, while loneliness, in turn, motivates in-
creased social media usage.
The study findings reinforce concerns that while
SNS can serve as a medium for social connection,
frequent or prolonged use might aggravate feelings of
isolation and impair sleep quality, impacting overall
well-being.
It is important to note that FCA turned out to be an
efficient and useful approach to find aspects not easily
identified at first.
Future research could employ longitudinal data to
track changes over time, further validating the impact
of social media on mental health. Additionally, stud-
ies could incorporate diverse demographic datasets to
examine whether cultural differences influence these
behaviors.
ICEIS 2025 - 27th International Conference on Enterprise Information Systems
232

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
Alonzo, R., Hussain, J., Stranges, S., and Anderson, K. K.
(2021). Interplay between social media use, sleep
quality, and mental health in youth: A systematic re-
view. Sleep Medicine Reviews, 56:101414.
Buysse, D. J., Reynolds III, C. F., Monk, T. H., Berman,
S. R., and Kupfer, D. J. (1989). The pittsburgh sleep
quality index: a new instrument for psychiatric prac-
tice and research. Psychiatry research, 28(2):193–
213.
Carr, C. T. and Hayes, R. A. (2015). Social media: Defin-
ing, developing, and divining. Atlantic Journal of
Communication, 23(1):46–65.
Coutinho, L., Song, M., and Z
´
arate, L. (2024). Longitudi-
nal data analysis based on triadic rules to describe of
the psychological reactions during covid 19 pandemic.
In 17th International Joint Conference on Biomedi-
cal Engineering Systems and Technologies, volume 2,
pages 323–329.
Exelmans, L. and Van den Bulck, J. (2016). Bedtime mo-
bile phone use and sleep in adults. Social Science and
Medicine, 148:93–101.
Ganter, B. and Wille, R. (2012). Formal concept analysis:
mathematical foundations. Springer Science & Busi-
ness Media.
Hirshkowitz, M., Whiton, K., Albert, S. M., Alessi, C.,
Bruni, O., DonCarlos, L., Hazen, N., Herman, J.,
Katz, E. S., Kheirandish-Gozal, L., Neubauer, D. N.,
O’Donnell, A. E., Ohayon, M., Peever, J., Rawd-
ing, R., Sachdeva, R. C., Setters, B., Vitiello, M. V.,
Ware, J. C., and Adams Hillard, P. J. (2015). Na-
tional sleep foundation’s sleep time duration recom-
mendations: methodology and results summary. Sleep
Health, 1(1):40–43.
Islam, M. R., Tushar, M. I., Jannath, S., Moona, A. A., Ak-
ter, S., and Islam, S. M. A. (2021). Data set concerning
the use of social networking sites and mental health
problems among the young generation in bangladesh.
Data in Brief, 39:107593.
Kumar, P. M., Udaya Kumar, V., Meenakshi, S., Bahekar,
T. N., Narapaka, P. K., Dhingra, S., and Murti, K.
(2024). Chapter 1 - epidemiology and risk factors of
mental disorders. In Bagchi, D., Chaurasia, R. N., and
Ohia, S. E., editors, A Review on Diverse Neurological
Disorders, pages 3–12. Academic Press.
Kuss, D. J. and Griffiths, M. D. (2017). Social network-
ing sites and addiction: Ten lessons learned. Interna-
tional Journal of Environmental Research and Public
Health, 14(3).
Lana, P., Nobre, C., Zarate, L., and Song, M. (2022). For-
mal concept analysis applied to a longitudinal study
of covid-19. Scitepress.
Lehmann, F. and Wille, R. (1995). A triadic approach
to formal concept analysis. In Conceptual Struc-
tures: Applications, Implementation and Theory:
Third International Conference on Conceptual Struc-
tures, ICCS’95 Santa Cruz, CA, USA, August 14–18,
1995 Proceedings 3, pages 32–43. Springer.
Marttila, E., Koivula, A., and R
¨
as
¨
anen, P. (2021). Does
excessive social media use decrease subjective well-
being? a longitudinal analysis of the relationship be-
tween problematic use, loneliness and life satisfaction.
Telematics and Informatics, 59:101556.
Missaoui, R. and Emamirad, K. (2017). Lattice miner-a
formal concept analysis tool. In 14th International
Conference on Formal Concept Analysis, volume 3,
pages 2–5.
Orzeł-Gryglewska, J. (2010). Consequences of sleep de-
privation. International Journal of Occupational
Medicine and Environmental Health, 23(1):95 – 114.
O’Day, E. B. and Heimberg, R. G. (2021). Social media use,
social anxiety, and loneliness: A systematic review.
Computers in Human Behavior Reports, 3:100070.
Poelmans, J., Ignatov, D. I., Kuznetsov, S. O., and Dedene,
G. (2013). Formal concept analysis in knowledge pro-
cessing: A survey on applications. Expert Systems
with Applications, 40(16):6538–6560.
Russell, D. W. (1996). Ucla loneliness scale (version 3):
Reliability, validity, and factor structure. Journal of
personality assessment, 66(1):20–40.
Song, H., Zmyslinski-Seelig, A., Kim, J., Drent, A., Victor,
A., Omori, K., and Allen, M. (2014). Does facebook
make you lonely?: A meta analysis. Computers in
Human Behavior, 36:446–452.
Song, M. A. J., Galvez, L. E. Z., and Santos, D. S. (2024).
Applying formal concept analysis to characterize in-
fant mortality: Aplicando an
´
alise de formal de con-
ceitos para a caracterizac¸
˜
ao da mortalidade infantil.
Concilium, 24(3):387–407.
Wille, R. (1995). The basic theorem of triadic concept anal-
ysis. Order, 12:149–158.
Wille, R. (2009). Restructuring lattice theory: An approach
based on hierarchies of concepts. In Proceedings of
the 7th International Conference on Formal Concept
Analysis, ICFCA ’09, page 314–339, Berlin, Heidel-
berg. Springer-Verlag.
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.
Exploring Links Between Social Media Habits, Loneliness, and Sleep: A Formal Concept Analysis Approach
233