Journal of Engineering Design,
Год журнала:
2024,
Номер
unknown, С. 1 - 21
Опубликована: Апрель 12, 2024
Generally,
the
existing
methods
for
constructing
a
knowledge
graph
used
in
question
answering
system
adopted
two
different
models
respectively,
one
is
identifying
entities,
and
other
extracting
relationships
between
entities.
However,
this
method
may
reduce
quality
of
because
it
very
difficult
to
keep
contextual
information
consistent
with
same
entities
models.
To
address
issue,
paper
proposes
model
called
GPB
(GlobalPointer
+
BiLSTM)
which
integrates
BiLSTM
into
GlobalPointer
through
concatenation
operations
simultaneously
guarantee
rationality
identified
In
addition,
enhance
user
experience
using
an
intelligent
motor
fault
maintenance
system,
BAC
(BiLSTM
Attention
CRF)
proposed
identify
named
questions,
BERT-wwm
classify
intentions
improve
answers.
Finally,
verify
advantages
BAC,
comparative
experiments
real
application
effects
developed
are
demonstrated
on
our
built
dataset.
The
experimental
results
indicate
that
constructed
provide
engineers
high-quality
services.
Journal of Engineering Design,
Год журнала:
2024,
Номер
unknown, С. 1 - 23
Опубликована: Май 20, 2024
Machine
learning
has
been
widely
used
in
design
activities,
enabling
more
informed
decision-making.
However,
high-performance
machine
models,
often
referred
to
as
'black-box',
result
a
lack
of
explainability
regarding
predictions.
The
absence
erodes
the
trust
between
designers
and
these
models
hinders
human-machine
collaboration
for
desirable
decisions.
Explainable
AI
focuses
on
creating
explanations
that
are
accessible
comprehensible
stakeholders,
thereby
improving
explainability.
A
recent
advancement
field
explainable
involves
leveraging
domain-specific
knowledge
via
graph.
Additionally,
advent
large
language
like
ChatGPT,
acclaimed
their
ability
output
domain
knowledge,
perform
complex
processing,
support
seamless
end-user
interaction,
potential
expand
horizons
AI.
Inspired
by
developments,
we
propose
novel
hybrid
method
synergizes
ChatGPT
graph
augment
post-hoc
context.
outcome
is
generation
contextual
meaningful
explanations,
with
added
possibility
further
interaction
uncover
deeper
insights.
effectiveness
proposed
illustrated
through
case
study
customer
segmentation.
