2022 17th Iberian Conference on Information Systems and Technologies (CISTI),
Год журнала:
2023,
Номер
2, С. 1 - 6
Опубликована: Июнь 20, 2023
The
importance
of
education
has
given
rise
to
the
search
for
different
learning
techniques
and
use
information
technology
tools.
One
currently
used
achieve
encourage
is
serious
games
gamification;
these
allow
students
objectives
while
playing.
This
paper
presents
a
systematic
literature
review
about
proposals
web
programming
based
on
games;
10
studies
are
presented
content
related
programming,
languages,
topics
addressed
in
teaching.
study
contributes
who
want
learn
programming.
Large
language
models
(LLMs)
have
recently
been
applied
in
software
engineering
to
perform
tasks
such
as
translating
code
between
programming
languages,
generating
from
natural
language,
and
autocompleting
it
is
being
written.
When
used
within
development
tools,
these
systems
typically
treat
each
model
invocation
independently
all
previous
invocations,
only
a
specific
limited
functionality
exposed
the
user
interface.
This
approach
interaction
misses
an
opportunity
for
users
more
deeply
engage
with
by
having
context
of
their
interactions,
well
code,
inform
model's
responses.
We
developed
prototype
system
–
Programmer's
Assistant
order
explore
utility
conversational
interactions
grounded
engineers'
receptiveness
idea
conversing
with,
rather
than
invoking,
code-fluent
LLM.
Through
evaluation
42
participants
varied
levels
experience,
we
found
that
our
was
capable
conducting
extended,
multi-turn
discussions,
enabled
additional
knowledge
capabilities
beyond
generation
emerge
Despite
skeptical
initial
expectations
assistance,
were
impressed
breadth
assistant's
capabilities,
quality
its
responses,
potential
improving
productivity.
Our
work
demonstrates
unique
LLMs
co-creative
processes
like
development.
Large
Language
Models
(LLMs)
have
the
potential
to
fundamentally
change
way
people
engage
in
computer
programming.
Agent-based
modeling
(ABM)
has
become
ubiquitous
natural
and
social
sciences
education,
yet
no
prior
studies
explored
of
LLMs
assist
it.
We
designed
NetLogo
Chat
support
learning
practice
NetLogo,
a
programming
language
for
ABM.
To
understand
how
users
perceive,
use,
need
LLM-based
interfaces,
we
interviewed
30
participants
from
global
academia,
industry,
graduate
schools.
Experts
reported
more
perceived
benefits
than
novices
were
inclined
adopt
their
workflow.
found
significant
differences
between
experts
perceptions,
behaviors,
needs
human-AI
collaboration.
surfaced
knowledge
gap
as
possible
reason
benefit
gap.
identified
guidance,
personalization,
integration
major
interfaces
Existing
research
on
human-AI
collaborative
decision-making
focuses
mainly
the
interaction
between
AI
and
individual
decision-makers.
There
is
a
limited
understanding
of
how
may
perform
in
group
decision-making.
This
paper
presents
wizard-of-oz
study
which
two
participants
an
form
committee
to
rank
three
English
essays.
One
novelty
our
that
we
adopt
speculative
design
by
endowing
equal
power
humans
We
enable
discuss
vote
equally
with
other
human
members.
find
although
voice
considered
valuable,
still
plays
secondary
role
because
it
cannot
fully
follow
dynamics
discussion
make
progressive
contributions.
Moreover,
divergent
opinions
regarding
"equal
AI"
shed
light
possible
future
relations.
ACM Transactions on Software Engineering and Methodology,
Год журнала:
2024,
Номер
unknown
Опубликована: Окт. 7, 2024
Remote
pair
programming
is
widely
used
in
software
development,
but
no
research
has
examined
how
race
affects
these
interactions
between
developers.
We
embarked
on
this
study
due
to
the
historical
under
representation
of
Black
developers
tech
industry,
with
White
comprising
majority.
Our
involved
24
experienced
developers,
forming
12
gender-balanced
same-
and
mixed-race
pairs.
Pairs
collaborated
a
task
using
think-aloud
method,
followed
by
individual
retrospective
interviews.
findings
revealed
elevated
productivity
scores
for
pairs,
differences
code
quality
Mixed-race
pairs
excelled
distribution,
shared
decision-making,
role-exchange
encountered
communication
challenges,
discomfort,
anxiety,
shedding
light
complexity
diversity
dynamics.
emphasizes
race’s
impact
remote
underscores
need
diverse
tools
methods
address
racial
disparities
collaboration.
Proceedings of the 30th ACM Joint European Software Engineering Conference and Symposium on the Foundations of Software Engineering,
Год журнала:
2022,
Номер
unknown, С. 319 - 331
Опубликована: Ноя. 7, 2022
Recent
research
has
shown
feasibility
of
an
interactive
pair-programming
conversational
agent,
but
implementing
such
agent
poses
three
challenges:
a
lack
benchmark
datasets,
absence
software
engineering
specific
labels,
and
the
need
to
understand
developer
conversations.
To
address
these
challenges,
we
conducted
Wizard
Oz
study
with
14
participants
pair
programming
simulated
collected
4,443
developer-agent
utterances.
Based
on
this
dataset,
created
26
labels
using
open
coding
process
develop
hierarchical
classification
scheme.
labeled
conversations,
compared
accuracy
state-of-the-art
transformer-based
language
models,
BERT,
GPT-2,
XLNet,
which
performed
interchangeably.
In
order
begin
creating
researchers
practitioners
conduct
resource
intensive
studies.
Presently,
there
exists
vast
amounts
developer-developer
conversations
video
hosting
websites.
investigate
publicly
available
dataset
(3,436
utterances)
our
scheme
found
that
BERT
model
trained
data
~10%
worse
than
data,
when
transfer-learning,
improved.
Finally,
qualitative
analysis
revealed
are
more
implicit,
neutral,
opinionated
Our
results
have
implications
for
developing
agents.
