Comparative Analysis with Intelligent Visual Interfaces
Takanori Fujiwara
Department of Computer Science
University of California, Davis, CA 95616, USA
Via https://zoom.us/j/93142929445?pwd=bytvSk1pN3hwbEJXZS9oVEUwcnRSZz09
Abstract
Comparison—the
act
of
finding
similarities
and
differences
between
two
or
more
groups
within
datasets—is
rooted
as
a
fundamental
analysis
task.
However,
this
task
is
non-trivial
when
analyzing
large
network
or
high-dimensional
datasets.
In
such
cases,
it
is
difficult
to
identify
the
key
contributing
factors
to
the
similarities
and
differences
of
different
groups
from
all
possible
relationships
or
attributes.
Representation
learning
can
help
address
this
process
by
extracting
influential
factors
for
a
particular
aspect
within
a
dataset
(e.g.,
data
variance).
But,
due
to
their
inability
to
cover
a
wide
range
of
data
types
and
analysis
targets,
these
existing
techniques
have
limited
capabilities
for
comparative
analysis.
In
this
talk,
I
will
address
the
challenges
of
comparative
analysis
with
intelligent
visual
interfaces
that
couple
interactive
visualizations
and
contrastive
learning—a
new
emerging
representation
learning
scheme
that
finds
salient
patterns
in
one
dataset
relative
to
another
dataset.
I
will
demonstrate
the
effectiveness
of
these
intelligent
visual
interfaces
for
network
data
and
high-dimensional
data
comparisons
by
analyzing
real-world
datasets.
Finally,
I
will
discuss
my
future
plans
to
develop
a
new
interactive
representing
learning
method
in
addition
to
future
research
directions
that
will
further
expand
the
field
of
comparative
analysis.
Biographical Sketch
Takanori Fujiwara is a Ph.D. candidate at the Department of Computer Science at the University of California, Davis where he is a member of the Visualization and Interface Design Innovation research group, advised by Dr. Kwan‑Liu Ma. He works at the intersection of data science and data visualization where his current research focuses on developing techniques in visual analytics, machine learning, and network science to analyze high-dimensional and network data. He is especially interested in how the combination of representation learning and interactive visualization can aid comparative analysis. He has published his research in top-tier visualization venues at the IEEE Transactions on Visualization and Computer Graphics and the IEEE VIS conferences. His work received a Best Paper Honorable Mention at the IEEE VIS in 2019 and the Best Graduate Researcher Award from the Department of Computer Science at UC Davis in 2020. Before UC Davis, he received his Master's degree in Environmental Science and B.E. in Systems Innovation from the University of Tokyo and has worked for Kajima Corporation in Japan.