Health Policy,
Год журнала:
2023,
Номер
132, С. 104821 - 104821
Опубликована: Апрель 14, 2023
Increasingly,
the
credibility
of
online
reviews
is
drawing
critical
attention
due
to
lack
control
mechanisms,
constant
debate
about
fake
and,
last
but
not
least,
current
developments
in
field
artificial
intelligence.
For
this
reason,
aim
study
was
examine
extent
which
assessments
recorded
on
physician
rating
websites
(PRWs)
are
credible,
based
a
comparison
other
evaluation
criteria.Referring
PRISMA
guidelines,
comprehensive
literature
search
conducted
across
different
scientific
databases.
Data
were
synthesized
by
comparing
individual
statistical
outcomes,
objectives
and
conclusions.The
chosen
strategy
led
database
36,755
studies
28
ultimately
included
systematic
review.
The
review
yielded
mixed
results
regarding
PRWs.
While
seven
publications
supported
PRWs,
six
found
no
correlation
between
PRWs
alternative
datasets.
15
reported
results.This
has
shown
that
ratings
seem
be
credible
when
relying
primarily
patients'
perception.
However,
these
portals
inadequate
represent
comparative
values
such
as
medical
quality
physicians.
health
policy
makers
our
show
decisions
perceptions
may
well
data
from
all
decisions,
however,
do
contain
sufficiently
useful
data.
Mathematics,
Год журнала:
2024,
Номер
12(4), С. 566 - 566
Опубликована: Фев. 13, 2024
Currently,
direct
surveys
are
used
less
and
to
assess
satisfaction
with
the
quality
of
user
services.
One
most
effective
methods
solve
this
problem
is
extract
attitudes
from
social
media
texts
using
natural
language
text
mining.
This
approach
helps
obtain
more
objective
results
by
increasing
representativeness
independence
sample
service
consumers
being
studied.
The
purpose
article
improve
existing
test
a
method
for
classifying
Russian-language
reviews
patients
about
work
medical
institutions
doctors,
extracted
resources.
authors
developed
hybrid
tested
machine
learning
various
neural
network
architectures
(GRU,
LSTM,
CNN)
achieve
goal.
More
than
60,000
posted
on
two
popular
doctor
review
sites
in
Russia
were
analysed.
Main
results:
(1)
classification
algorithm
highly
efficient—the
best
result
was
shown
GRU-based
architecture
(val_accuracy
=
0.9271);
(2)
application
searching
named
entities
messages
after
their
division
made
it
possible
increase
efficiency
each
classifiers
based
use
artificial
networks.
study
has
scientific
novelty
practical
significance
field
demographic
research.
To
classification,
future,
planned
expand
semantic
object
appeal
sentiment
take
into
account
resulting
fragments
separately
other.
Journal of Decision System,
Год журнала:
2024,
Номер
unknown, С. 1 - 33
Опубликована: Окт. 7, 2024
Generative
Artificial
Intelligence
(Gen-AI)
is
a
burgeoning
subfield
of
artificial
intelligence
that
focuses
on
creating
new
content
which
poised
to
revolutionise
different
industries
by
2028.
This
study
aims
first
identify
key
enablers
for
the
successful
integration
Gen-AI
into
supply
chain
with
help
Delphi
and
AHP
techniques.
Then,
we
screened
these
categories
identified
seven
enabler
using
method.
We
computed
weights
those
ranked
them
basis
their
Ethical
Fair
AI
Practices
Public
Trust
Societal
Impact
among
most
significant.
Second,
this
categorised
tweet
positive,
neutral,
negative
sentiments
sentiment
analysis
fifteen
topics
from
secondary
data.
The
research
concludes
actionable
strategies
practitioners
outlines
significance
ethical
trust-related
in
adoption
chain.
Health Policy,
Год журнала:
2023,
Номер
132, С. 104821 - 104821
Опубликована: Апрель 14, 2023
Increasingly,
the
credibility
of
online
reviews
is
drawing
critical
attention
due
to
lack
control
mechanisms,
constant
debate
about
fake
and,
last
but
not
least,
current
developments
in
field
artificial
intelligence.
For
this
reason,
aim
study
was
examine
extent
which
assessments
recorded
on
physician
rating
websites
(PRWs)
are
credible,
based
a
comparison
other
evaluation
criteria.Referring
PRISMA
guidelines,
comprehensive
literature
search
conducted
across
different
scientific
databases.
Data
were
synthesized
by
comparing
individual
statistical
outcomes,
objectives
and
conclusions.The
chosen
strategy
led
database
36,755
studies
28
ultimately
included
systematic
review.
The
review
yielded
mixed
results
regarding
PRWs.
While
seven
publications
supported
PRWs,
six
found
no
correlation
between
PRWs
alternative
datasets.
15
reported
results.This
has
shown
that
ratings
seem
be
credible
when
relying
primarily
patients'
perception.
However,
these
portals
inadequate
represent
comparative
values
such
as
medical
quality
physicians.
health
policy
makers
our
show
decisions
perceptions
may
well
data
from
all
decisions,
however,
do
contain
sufficiently
useful
data.
