International Journal of Consumer Studies,
Год журнала:
2025,
Номер
49(3)
Опубликована: Апрель 24, 2025
ABSTRACT
Artificial
intelligence
service
agent
(AISA)
anthropomorphism
is
increasingly
applied
in
situations
to
deliver
consumers
a
more
personalized
experience.
Extensive
research
has
explored
it
over
the
past
decade.
However,
body
of
knowledge
about
this
domain
remains
fragmented
due
lack
comprehensive
review.
This
paper
aims
understand
through
framework‐based
systematic
literature
review
approach.
We
conducted
an
in‐depth
analysis
149
peer‐reviewed
articles
retrieved
on
April
25,
2024.
Significant
results
include
(a)
analyzing
profiling
(i.e.,
publication,
journal,
citation,
and
country
analysis),
(b)
synthesizing
findings
based
theories‐contexts‐characteristics‐methods
(TCCM)
framework,
(c)
proposing
future
agendas
TCCM
framework.
The
present
concludes
that
experiencing
rapid
growth,
evidenced
by
increasing
number
citations,
wide
range
publications
reputable
journals,
notable
surge
international
collaboration.
Furthermore,
AISA
affects
outcomes
related
AISA,
product,
brand,
even
company
via
cognitive,
affective,
social
mediators.
Service‐,
AISA‐,
consumer‐related
factors
moderate
these
effects.
Based
gap
analysis,
we
propose
directions.
Our
advances
affecting
consumer
responses.
It
offers
valuable
insights
for
practitioners
effectively
deploying
anthropomorphic
AISAs
serve
appropriate
contexts.
Aslib Journal of Information Management,
Год журнала:
2025,
Номер
unknown
Опубликована: Апрель 11, 2025
Purpose
The
rapid
growth
of
live
streaming
has
intensified
competition
among
streamers.
A
key
challenge
lies
in
aligning
live-streaming
sales
strategies
with
consumer
needs
to
cultivate
positive
attitudes.
Design/methodology/approach
Drawing
on
theories
the
types
consumer–live
streamer
interaction
strategies,
uses
and
satisfaction
theory,
channel
complementarity
theory
congruence
hypothesis
content
online
reviews,
this
study
employs
quantitative
research
investigate
how
effectively
integrate
genuine
product
feature
preferences,
as
reflected
messages
conveyed
by
To
achieve
objective,
paper
synthesizes
machine
learning
techniques
–
including
information
quantity
calculation
BERT
statistical
analysis
methods
such
difference
testing
scenario
experiments.
Findings
results
demonstrate
that
proposed
predictive
model
perceived
useful
information,
which
relies
quantity,
provides
notable
advantages.
Specifically,
there
is
a
significant
disparity
between
preferences
derived
from
those
identified
through
traditional
consider
all
available
data.
Moreover,
extracted
diverge
focal
points
streamer’s
actual
explanations.
method,
aligns
explanations,
markedly
enhances
consumers’
purchase
intentions.
Originality/value
This
introduces
novel
metric
distinguish
reviews
broader
dataset,
integrating
it
sentiment
extraction
techniques.
Unlike
treat
equally
valuable,
approach
prioritizes
comments
based
informational
value,
emotional
tone
relevance
features,
providing
nuanced
precise
understanding
preferences.
By
these
streamers’
offers
data-driven,
consumer-centric
strategy
for
improving
recommendations
engagement
commerce.
comprehensive
framework
represents
advancement
over
models
reliant
data
volume
alone.
