The
catastrophic
earthquake
that
struck
Morocco
on
Septem-
ber
8,
2023,
garnered
significant
media
coverage,
leading
to
the
swift
dissemination
of
information
across
various
social
and
online
plat-
forms.
However,
heightened
visibility
also
gave
rise
a
surge
in
fake
news,
presenting
formidable
challenges
efficient
distribution
ac-
curate
crucial
for
effective
crisis
management.
This
paper
introduces
an
innovative
approach
detection
by
integrating
Natural
language
processing,
bidirectional
long-term
memory
(Bi-LSTM),
con-
volutional
neural
network
(CNN),
hierarchical
attention
(HAN)
models
within
context
this
seismic
event.
Leveraging
ad-
vanced
machine
learning,deep
learning,
data
analysis
techniques,
we
have
devised
sophisticated
news
model
capable
precisely
identifying
categorizing
misleading
information.
amal-
gamation
these
enhances
accuracy
efficiency
our
system,
addressing
pressing
need
reliable
amidst
chaos
crisis.
Wiley Interdisciplinary Reviews Data Mining and Knowledge Discovery,
Год журнала:
2025,
Номер
15(1)
Опубликована: Фев. 23, 2025
ABSTRACT
Social
media,
particularly
microblogging
platforms,
are
essential
for
rapid
information
sharing
and
public
discussion
but
often
allow
rumors,
that
is,
unverified
information,
to
spread
rapidly
during
events
or
persist
over
time.
These
platforms
also
offer
opportunities
study
the
dynamics
of
rumors
develop
computational
methods
assess
their
veracity.
In
this
paper,
we
provide
a
comprehensive
review
existing
theoretical
foundations,
interdisciplinary
challenges,
emerging
advancements
in
rumor
detection
research,
with
focus
on
integrating
approaches.
Drawing
insights
from
computer
science,
cognitive
psychology,
sociology,
explore
methodologies,
such
as
multimodal
fusion,
graph‐based
models,
attention
mechanisms,
while
highlighting
gaps
real‐world
scalability,
ethical
transparency,
cross‐platform
adaptability.
Using
systematic
literature
bibliometric
analysis,
identify
trends,
methods,
current
research.
Our
findings
emphasize
collaboration
adaptable,
efficient,
strategies.
We
highlight
critical
role
combining
socio‐psychological
advanced
techniques
address
human
factors
spread.
Furthermore,
importance
designing
systems
remain
effective
across
diverse
cultural
linguistic
contexts,
enhancing
global
applicability.
propose
conceptual
framework
theories
techniques,
offering
roadmap
improving
addressing
misinformation
challenges
platforms.
Information,
Год журнала:
2023,
Номер
14(10), С. 541 - 541
Опубликована: Окт. 3, 2023
Recent
developments
in
IoT,
big
data,
fog
and
edge
networks,
AI
technologies
have
had
a
profound
impact
on
number
of
industries,
including
medical.
The
use
for
therapeutic
purposes
has
been
hampered
by
its
inexplicability.
Explainable
Artificial
Intelligence
(XAI),
revolutionary
movement,
arisen
to
solve
this
constraint.
By
using
decision-making
prediction
outputs,
XAI
seeks
improve
the
explicability
standard
models.
In
study,
we
examined
global
empirical
research
medical
field.
bibliometric
analysis
tools
VOSviewer
Biblioshiny
were
used
examine
171
open
access
publications
from
Scopus
database
(2019–2022).
Our
findings
point
several
prospects
growth
area,
notably
areas
medicine
like
diagnostic
imaging.
With
109
articles
healthcare
classification,
prediction,
diagnosis,
USA
leads
world
output.
88
citations,
IEEE
Access
greatest
all
journals.
extensive
survey
covers
range
applications
healthcare,
such
as
therapy,
prevention,
palliation,
offers
helpful
insights
researchers
who
are
interested
This
report
provides
direction
future
industry
endeavors.
2022 10th International Conference on Reliability, Infocom Technologies and Optimization (Trends and Future Directions) (ICRITO),
Год журнала:
2024,
Номер
unknown, С. 1 - 6
Опубликована: Март 14, 2024
By
examining
three
research
topics,
this
review
paper
examines
the
state
of
blockchain
technology
research.
To
grasp
most
recent
improvements
and
innovations,
it
is
first
necessary
to
analyse
advancement
in
field
blockchain.
Second,
nations
organisations
that
have
significantly
contributed
study
are
named.
Finally,
field's
well-liked
areas
keyword
trends
examined.
The
study's
conclusions
provide
a
comprehensive
assessment
technol-ogy
at
present,
highlighting
significant
developments,
institutions,
new
fields
study.
This
analysis
will
help
scholars,
decision-makers,
business
experts
comprehend
current
knowledge
suggest
promising
for
collaboration
area
Applied Sciences,
Год журнала:
2023,
Номер
13(7), С. 4207 - 4207
Опубликована: Март 26, 2023
In
the
digital
age,
social
media
platforms
are
becoming
vital
tools
for
generating
and
detecting
deepfake
news
due
to
rapid
dissemination
of
information.
Unfortunately,
today,
fake
is
being
developed
at
an
accelerating
rate
that
can
cause
substantial
problems,
such
as
early
detection
news,
a
lack
labelled
data
available
training,
identifying
instances
still
need
be
discovered.
Identifying
false
requires
in-depth
understanding
authors,
entities,
connections
between
words
in
long
text.
many
deep
learning
(DL)
techniques
have
proven
ineffective
with
lengthy
texts
address
these
issues.
This
paper
proposes
TL-MVF
model
based
on
transfer
media.
To
generate
sentences,
T5,
or
Text-to-Text
Transfer
Transformer
model,
was
employed
cleaning
feature
extraction.
next
step,
we
designed
optimal
hyperparameter
RoBERTa
effectively
real
news.
Finally,
propose
multiplicative
vector
fusion
classifying
from
efficiently.
A
real-time
benchmarked
dataset
used
test
validate
proposed
model.
For
F-score,
accuracy,
precision,
recall,
AUC
were
performance
evaluation
measures.
As
result,
performed
better
than
existing
benchmarks.
Electronics,
Год журнала:
2023,
Номер
12(13), С. 2942 - 2942
Опубликована: Июль 4, 2023
With
the
spread
of
Internet
technologies,
use
social
media
has
increased
exponentially.
Although
many
benefits,
it
become
primary
source
disinformation
or
fake
news.
The
news
is
creating
societal
and
economic
issues.
It
very
critical
to
develop
an
effective
method
detect
so
that
can
be
stopped,
removed
flagged
before
spreading.
To
address
challenge
accurately
detecting
news,
this
paper
proposes
a
solution
called
Statistical
Word
Embedding
over
Linguistic
Features
via
Deep
Learning
(SWELDL
Fake),
which
utilizes
deep
learning
techniques
improve
accuracy.
proposed
model
implements
statistical
“principal
component
analysis”
(PCA)
on
textual
representations
identify
significant
features
help
In
addition,
word
embedding
employed
comprehend
linguistic
Bidirectional
Long
Short-Term
Memory
(Bi-LSTM)
utilized
classify
as
true
fake.
We
used
benchmark
dataset
SWELDL
Fake
validate
our
model,
about
72,000
articles
collected
from
different
datasets.
Our
achieved
classification
accuracy
98.52%
surpassing
performance
state-of-the-art
machine
models.
