Recent
advancements
in
large
language
models
(LLMs)
hold
significant
promise
for
improving
physics
education
research
that
uses
machine
learning.
In
this
study,
we
compare
the
application
of
various
conducting
a
large-scale
analysis
written
text
grounded
classification
problem:
identifying
skills
students’
typed
lab
notes
through
sentence-level
labeling.
Specifically,
use
training
data
to
fine-tune
two
different
LLMs,
BERT
and
LLaMA,
performance
these
both
traditional
bag-of-words
approach
few-shot
LLM
(without
fine-tuning).
We
evaluate
based
on
their
resource
use,
metrics,
outcomes
when
notes.
find
higher-resource
often,
but
not
necessarily,
perform
better
than
lower-resource
models.
also
all
report
similar
trends
outcomes,
although
absolute
values
estimated
measurements
are
always
within
uncertainties
each
other.
results
discuss
relevant
considerations
researchers
seeking
select
model
type
as
classifier.
Published
by
American
Physical
Society
2025
Research in Science Education,
Journal Year:
2024,
Volume and Issue:
54(5), P. 977 - 997
Published: June 27, 2024
Abstract
The
use
of
Artificial
Intelligence
(AI)
in
education
is
transforming
various
dimensions
the
system,
such
as
instructional
practices,
assessment
strategies,
and
administrative
processes.
It
also
plays
an
active
role
progression
science
education.
This
systematic
review
attempts
to
render
inherent
understanding
evidence-based
interaction
between
AI
Specifically,
this
study
offers
a
consolidated
analysis
AI’s
impact
on
students’
learning
outcomes,
contexts
its
adoption,
teachers’
perceptions
about
use,
challenges
within
present
followed
PRISMA
guidelines
empirical
papers
published
from
2014
2023.
In
total,
74
records
met
eligibility
for
study.
Previous
research
provides
evidence
integration
into
variety
fields
physical
natural
sciences
many
countries
across
globe.
results
revealed
that
AI-powered
tools
are
integrated
achieve
pedagogical
benefits,
including
enhancing
environment,
creating
quizzes,
assessing
work,
predicting
their
academic
performance.
findings
paper
have
implications
teachers,
educational
administrators,
policymakers.
Information,
Journal Year:
2024,
Volume and Issue:
15(2), P. 99 - 99
Published: Feb. 8, 2024
GPT
(Generative
Pre-trained
Transformer)
represents
advanced
language
models
that
have
significantly
reshaped
the
academic
writing
landscape.
These
sophisticated
offer
invaluable
support
throughout
all
phases
of
research
work,
facilitating
idea
generation,
enhancing
drafting
processes,
and
overcoming
challenges
like
writer’s
block.
Their
capabilities
extend
beyond
conventional
applications,
contributing
to
critical
analysis,
data
augmentation,
design,
thereby
elevating
efficiency
quality
scholarly
endeavors.
Strategically
narrowing
its
focus,
this
review
explores
alternative
dimensions
LLM
specifically
augmentation
generation
synthetic
for
research.
Employing
a
meticulous
examination
412
works,
it
distills
selection
77
contributions
addressing
three
questions:
(1)
on
Generating
Research
data,
(2)
Data
Analysis,
(3)
Design.
The
systematic
literature
adeptly
highlights
central
focus
encapsulating
48
pertinent
contributions,
extends
proactive
role
in
analysis
shaping
design.
Pioneering
comprehensive
classification
framework
“GPT’s
use
Data”,
study
classifies
existing
into
six
categories
14
sub-categories,
providing
profound
insights
multifaceted
applications
data.
This
meticulously
compares
54
pieces
literature,
evaluating
domains,
methodologies,
advantages
disadvantages,
scholars
with
crucial
seamless
integration
across
diverse
their
pursuits.
Frontiers in artificial intelligence and applications,
Journal Year:
2024,
Volume and Issue:
unknown
Published: June 5, 2024
The
general
availability
of
large
language
models
and
thus
unrestricted
usage
in
sensitive
areas
everyday
life,
such
as
education,
remains
a
major
debate.
We
argue
that
employing
generative
artificial
intelligence
(AI)
tools
warrants
informed
examined
their
impact
on
problem
solving
strategies
higher
education.
In
study,
students
with
background
physics
were
assigned
to
solve
exercises,
one
group
having
access
an
internet
search
engine
(N=12)
the
other
being
allowed
use
ChatGPT
(N=27).
evaluated
performance,
strategies,
interaction
provided
tools.
Our
results
showed
nearly
half
solutions
support
mistakenly
assumed
be
correct
by
students,
indicating
they
overly
trusted
even
field
expertise.
Likewise,
42%
cases,
used
copy
&
paste
query
—
approach
only
4%
queries
highlighting
stark
differences
behavior
between
groups
limited
task
reflection
when
using
ChatGPT.
our
work,
we
demonstrated
need
(1)
guide
how
interact
LLMs
(2)
create
awareness
potential
shortcomings
for
users.
The
well-known
artificial
intelligence-based
chatbot
ChatGPT-4
has
become
able
to
process
image
data
as
input
in
October
2023.
We
investigated
its
performance
on
the
test
of
understanding
graphs
kinematics
inform
physics
education
community
current
potential
using
ChatGPT
process,
particularly
tasks
that
involve
graphical
interpretation.
found
ChatGPT,
average,
performed
similarly
students
taking
a
high-school
level
course,
but
with
important
differences
distribution
correctness
responses,
well
terms
displayed
“reasoning”
and
“visual”
abilities.
While
was
very
successful
at
proposing
productive
strategies
for
solving
expressed
correct
reasoning
most
it
had
difficulties
correctly
“seeing”
graphs.
suggest
that,
based
performance,
caution
critical
approach
are
needed
if
one
intends
use
role
tutor,
model
student,
or
tool
assisting
vision-impaired
persons
context
Published
by
American
Physical
Society
2024
Human Behavior and Emerging Technologies,
Journal Year:
2024,
Volume and Issue:
2024, P. 1 - 20
Published: Feb. 22, 2024
The
increasing
utilization
of
virtual
teams—driven
by
advancements
in
information
and
communication
technology
the
forces
globalization—has
spurred
significant
growth
both
theoretical
empirical
research.
Based
on
smart
literature
review
framework,
this
study
harnesses
artificial
intelligence
techniques,
specifically
natural
language
processing
topic
modeling,
to
extensively
analyze
trends
team
research
spanning
last
four
decades.
Analyses
a
dataset
comprising
2,184
articles
from
Scopus-indexed
journals
discern
16
distinct
topics,
encompassing
critical
areas
such
as
communication,
leadership,
trust.
trajectory
topics
field
has
witnessed
diversification
over
time.
Key
subjects
learning,
trust,
leadership
have
consistently
maintained
their
presence
among
ten
most
frequently
explored
topics.
In
contrast,
emerging
agile
development
patient
care
recently
become
some
prominent
themes.
Employing
state-of-the-art
modeling
technique,
BERTopic,
furnishes
comprehensive
dynamic
panorama
evolving
landscape
within
International Journal of Mathematical Education in Science and Technology,
Journal Year:
2024,
Volume and Issue:
unknown, P. 1 - 30
Published: May 28, 2024
This
study
investigates
the
utilisation
and
perceptions
of
artificial
intelligence
(AI)
applications
among
mathematics
science
teachers
in
enhancing
students'
learning
experience
classrooms.
One
prominent
AI
application
this
research
seeks
to
explore,
given
its
recent
rise
popularity
usage,
is
ChatGPT.
The
aims
explore
frequency
purpose
ChatGPT
potential
for
student
engagement,
implementation's
perceived
effectiveness,
benefits,
challenges.
A
sample
109
respondents
was
surveyed
gather
insights
into
usage
patterns
findings
shed
light
on
how
utilise
ChatGPT,
including
generating
examples,
assessing
difficulty,
providing
explanations,
supporting
problem-solving,
test
preparation.
Additionally,
reveals
that
shows
promise
engagement
understanding.
However,
highlighted
challenges
such
as
accuracy,
connectivity,
language
limitations.
These
contribute
our
understanding
role
technologies,
specifically
educational
settings
provide
valuable
effective
integration
teaching
practices
various
course
contents
topics.
Trends in Higher Education,
Journal Year:
2025,
Volume and Issue:
4(1), P. 6 - 6
Published: Jan. 27, 2025
Most
of
today’s
educators
are
in
no
shortage
digital
and
online
learning
technologies
available
at
their
fingertips,
ranging
from
Learning
Management
Systems
such
as
Canvas,
Blackboard,
or
Moodle,
meeting
tools,
homework,
tutoring
systems,
exam
proctoring
platforms,
computer
simulations,
even
virtual
reality/augmented
reality
technologies.
Furthermore,
with
the
rapid
development
wide
availability
generative
artificial
intelligence
(GenAI)
services
ChatGPT,
we
just
beginning
harnessing
potential
to
transform
higher
education.
Yet,
facing
large
number
options
provided
by
cutting-edge
technology,
an
imminent
question
on
mind
most
is
following:
how
should
I
choose
integrate
them
into
my
teaching
process
so
that
they
would
best
support
student
learning?
We
contemplate
over
these
types
important
timely
questions
share
our
reflections
evidence-based
approaches
tools
using
a
Self-regulated
Engaged
Framework
have
employed
research
physics
education
can
be
valuable
for
other
disciplines.
The
modern
physics
course
is
a
crucial
gateway
for
majors
as
it
provides
an
introduction
to
concepts
beyond
the
scope
of
K-12
education.
This
study
collected
167
syllabi
from
127
U.S.
research-intensive
institutions
using
publicly
available
(51.5%)
and
private
correspondence
(48.5%).
Due
size
corpus
being
large
enough
natural
language
processing
(NLP)
provide
significant
insight
into
curricular
content
course,
but
still
small
human
analysis
remain
feasible,
both
NLP
human-analysis
methods
were
utilized.
use
provided
way
test
newly
introduced
method
in
education
research
offered
novel
tool
cross-validation.
Public
catalogs
consulted
identify
prerequisites
corequisites
with
89.3%
students
having
completed
calculus
II
or
higher.
Foundational
topics
like
Newtonian
mechanics
(94%),
electricity
magnetism
(84.4%),
waves
optics
(77.2%)
frequently
required.
Most
consistently
taught
quantum
atomic
(83%),
relativity
(70%),
there
was
variability
inclusion
other
alongside
these
three
main
topics.
highlights
current
curricula
at
institutions,
emphasizing
importance
consistent
comprehensive
across
universities.
contributes
ongoing
discourse
on
optimizing
higher
Published
by
American
Physical
Society
2025
