►
Description
Presentation: December 3rd, 2021
Big Data Hubs Data Sharing and Cyberinfrastructure WG
Moderator: Melissa Cragin
Presenters: Kenton McHenry and Amanda Charbonneau
Title: EarthCube GeoCODES & Operationalizing FAIR at the Common Fund (Data Ecosystem)
Bios:
Kenton McHenry is the Associate Director for Software at the National Center for Supercomputing Applications (NCSA) at the University of Illinois at Urbana-Champaign. For the past 12 years Kenton has helped establish a team of Research Software Engineers to support the scientific community broadly in the software development, data management, and data analytics needs present in current research activities. Kenton has further served to drive the advancement of sustainable and reusable research cyberinfrastructure serving as a principal or co-principal investigator on efforts such as DIBBs Brown Dog to prototype a data transformation service, CSSI Clowder to support flexible data management leveraging machine learning, the Open Storage Network to pilot a national data fabric, PEcAn to drive infrastructure in support of ecological forecasting, and EarthCube piloting out activities to elevate software as full scholarly, peer reviewed, published objects.
Amanda Charbonneau is the Inreach and Outreach Coordinator for the Common Fund Data Ecosystem Coordination Center, an NIH initiative to increase access and use of NIH Common Fund datasets. She has an eclectic scientific background and has variously worked on topics such as population genetics, spectrometry, entomology, genomics, and building tank armor. She uses these experiences to train scientists to use new computational tools and data types, as well as to build and manage communities. Her interests include statistics, coding, video games, and occasionally tweeting at https://twitter.com/procrastinomics
A
A
She
has
quite
a
broad
and
eclectic
scientific
background
and
has
worked
previously
with
communities
like
population
genetics,
spectrometry,
entomology,
genomics
and
and
building
tank
armor.
I
love
that
she
uses
these
experiences
to
train
scientists
to
use
new
computational
tools
and
data
types
as
well
as
to
build
and
manage
communities.
Her
interests
include
statistics,
coding,
video
games
and
occasionally
tweeting,
and
I
can
post
that
link
a
bit
later
so
amanda
I'll
turn
it
over
to
you.
Thank
you.
A
B
B
But
I
always
like
to
start
this
kind
of
presentation
with
backing
up
and
talking
about
like
what
even
is
the
common
fund,
because
it's
not
a
thing
very
many
people
are
familiar
with,
so
you
probably
are
aware
that
the
nih
is
made
up
of
27
different
institutes
and
centers
and
that
each
of
one
of
those
has
its
own
like
deal
that
it
works
with.
So
the
national
cancer
institute
they
fund
stuff
about
cancer,
the
institute
of
allergy
infects
diseases,
they
fund
stuff
about
allergies
and
infectious
diseases,
the
eye
institute,
funds,
eye
stuff,
etc,
etc.
B
The
ability
to
do
genomic
sequencing
at
a
relatively
large
scale
started
to
really
take
off,
and
what
happened
at
nih
is
that
it
became
really
difficult
for
a
researcher
to
figure
out
when
they
had
a
new
cross-cutting
thing
that
they
wanted
to
do
like
they
wanted
to
do
genomics
on
allergies
who
to
send
that
funding
request
to
and
at
the
nih
when
they
got
funding.
Requests
like
this.
What
mostly
happened
was
the
allergy
institute
would
say
no,
no,
the
genomics
people
should
fund
this
and
the
genomics
people
would
say
no.
B
No,
the
allergy
people
should
fund
this,
and
so
things
weren't
getting
funded.
Nobody
really
know
what
knew
what
to
do.
So.
The
upshot
is
that
in
2006,
congress
created
a
new
ic
and
it's
called
the
common
fund.
It's
called
the
common
fund
because
they
literally
took
two
percent.
I
think
it
was
of
each
of
those
27
ics
budgets
pooled
it
in
a
common
bucket
and
said
this
is
going
to
be
what
funds
things
that
are
these
cross-cutting.
B
Nobody
can
agree
who
should
fund
its
stuff
so
in
order
to
be
funded
by
the
common
fund
you're
supposed
to
meet
these
five
criteria,
it's
for
transformative,
catalytic
synergistic,
cross-cutting,
unique
things
which
basically
is
just
a
whole
bunch
of
fancy
words
for
saying
stuff
that
fits
in
more
than
one.
I
see
that
everyone
was
arguing
about,
so
these
common
fund
programs
are
supposed
to
be
developing.
These
innovative
cross-cutting
data
sets
for
broad
use
that
wouldn't
be
funded
by
any
individual.
B
I
see
so
there
are
since
2006,
there
have
been
about
45,
funded
common
fund
programs
and
common
fund
doesn't
fund
individual
like
one
lab
science
projects.
What
they
fund
is
large
consortia,
so
the
one
that
almost
everyone
has
heard
of
is
the
human
microbiome
project,
so
the
human
microbiome
project
they
took
something
like
3,
000
people.
They
took
samples
from
stool.
They
tried
to
characterize
what
the
general
human
condition
of
of
intestinal
bacterial
communities
is.
B
They
also
did
some
studies
later
on
some
disease
states,
but
it
was
this
huge,
cross-cutting
study
and,
as
you
can,
as
you
can
imagine,
it
needed
people
with
familiarity
and
genomics
with
human
health
with
microbiomes
with
all
of
these
different
things.
So
it's
these
big,
cross-cutting
things,
another
one
that
a
lot
of
people
have
heard
of
is
gtex
and
what
they
do
is
they
are
trying
to
make
the
baseline
rna
level
map
of
every
human
tissue.
So
they
take
very
recently
deceased
cadavers.
B
They
take
samples
from
each
of
53
tissues,
they
run
rna-seq
analysis
on
them
and
they
are
trying
to
build
a
catalog
of
what
normal
rna-seq
levels
are
for
every
gene
for
all
of
these
different
tissues
and
I'm
not
going
to
go
through
every
single
one
of
these.
But
all
of
these
different
programs
are
these,
like
huge
scope,
trying
to
answer
these
cross-cutting
questions.
B
Now
there
are
some
issues
with
the
way
that
these
things
are
funded
and
that
they
have
evolved
over
time
so
congress
when
they
enacted
taking
a
little
bit
of
the
budget
from
each
of
these
places
and
dumping
it
in
the
middle.
They
said
that
when
you
fund
something
through
the
common
fund,
you
can't
fund
it
for
more
than
10
years
and
that
after
those
10
years,
somebody
else
needs
to
pick
up
the
funding.
But
at
no
point
did
anyone
figure
out
who
or
what
that
would
be.
B
They
kind
of
hoped
that
an
individual
ic
would
pick
up
the
funding
on
important
things,
but
that
actually
has
run
into
the
same
problem
as
the
original
funding,
which
is
that
all
of
the
different
ics
think
that
a
different
icc
should
pick
it
up.
So
there's
sort
of
a
sustainability
issue
with
common
fund
data
sets,
and
these
data
sets
were
all
this
transformative.
Catalytic
synergistic
cross-cutting,
unique
things.
B
B
So
that
is
why
the
common
fun
data
ecosystem
now
exists
our
job.
My
job
is
to
try
to
fix
this
problem,
to
increase
data
reuse
by
essentially
making
these
data
sets
interoperable
and
making
rules
as
the
wrong
word
best
practices
that,
as
new
common
fund
programs
are
funded,
they
can
use
so
that
interoperability
is
a
thing
that
happens
naturally,
across
data
sets
going
into
the
future,
rather
than
something
we
sort
of
have
to
back
track
and
fill
in
now.
B
B
So
real
quickly,
is
this
even
a
useful
thing
to
do
so?
You
could
imagine
two
different
ways
to
sort
of
jam
data
together
and
I
have
named
them
munging
and
layering,
because
that
is
how
it
works
in
my
brain.
So
munging
is
the
idea
of
there's
some
specific
rna-seq
data
set.
There's
some
other
specific
rna-seq
data
set.
They
have
a
whole
bunch
of
things
in
common
and
I
just
want
to
put
them
together
and
treat
them
as
one
really
big
rna-seq
data
set
so
that
I
can
answer
some
kind
of
question
right.
B
I
want
to
take
those
two
data
sets
munch
them
together,
and
this
can
be
really
really
difficult,
because
the
data
sets
either
have
to
have
a
ton
of
metadata
described
so
that
you
can
statistically
account
for
the
differences
between
them
or
that
are
not
interesting
like
blocking
effects
or
they
need
to
be
like
basically
exactly
the
same
as
one
another.
So
mundane
can
be
really
difficult,
but
you
could
imagine
other
ways
to
recombine
data,
which
would
be
something
like.
B
B
So
this
is
all
of
those
dc
all
of
those
common
fun
programs
that
I
had
the
logos
of
each
one
of
these
is
a
common
fund
program,
and
then
this
is
a
list
of
about
80
different
tissues,
and
so,
if
they
have
a
black
line
at
that
tissue,
they
have
data
on
that
tissue
and
if
they
don't,
then
they
don't
have
data
on
that
tissue.
So
you
can
see
that
without
reading
what
every
single
tissue
is
that
there
is
quite
a
lot
of
overlap
by
tissue
for
different
common
fund
programs.
B
So
we
probably
could
support
something
like
that
layering
we
could
get
different
or
sorry.
We
could
probably
support
something
like
the
munging
right
if
there's
the
same
kind
of
data
from
links
and
gtex
and
idg
and
comp
all
about
the
same
tissue.
Maybe
we
can
smush
data
together
that
way
and
we
also
looked
at
other
sorts
of
metadata,
so
this
is
sort
of
where
the
data
came
from.
There's
also
quite
a
lot
of
overlap
between
human
rights,
so
all
of
these
tissues
came
from
humans,
so
that
suggests
that
we
could
munch
things
together.
B
There's
disease
versus
healthy.
So
we
have
contrasts
that
we
could
compare
against
and
again
we
have
a
lot
of
this
across
all
of
these
different
programs,
and
then
we
asked
well
what,
if
we
don't
have
enough
similarity
there,
what
if
we
just
wanted
to
sort
of
stack
data
on
top
of
each
other,
so
this
top
one
is
showing
different
kinds
of
analyses
that
happened.
B
So
you
can
see
that
lynx
has
done
analyses
with
rna-seq
they've
done
metabolites,
binding,
epigenetic,
gross,
imaging
subcellular
imaging
and
protein
structure
and
gtex
has
whole
genome
sequencing
and
rna-seq
and
gross
imaging,
but
they
have
a
ton
of
overlap
in
tissues.
So,
probably,
if
I
went
and
found
the
tissue
that
I
care
about
at
gtex,
I
could
have
not
just
whole
genome
sequencing
and
rna-seq
and
gross
imaging
of
that,
but
maybe
I
could
find
corresponding
data
about
the
metabolites
or
about
the
epigenetic
information
or
subcellular
interesting,
imaging
or
other
things
like
that.
So
amanda.
B
C
Right
and
you
may
want
to
answer
this
later,
but
watching
and
listening
to
this,
I
wanted
to
ask
you
if
you
think
well,
first,
I
wanted
to
ask
you
how
many
pis
are
using
this
or
if
this
is
done
at
a
different
level,
and
then
I
wanted
to
bring
up
that.
One
of
the
the
projects
that
we
have
is
the
hubs.
C
You
know
the
metadata
for
all
the
awards,
the
thousands
of
awards
that
we'll
have
in
there,
but
it
actually
can
crawl.
Some
data
sets
too
and
find
some
stuff,
so
I'm
thinking
your
stuff
is
just
so
uber
cool
that
it
would
be
cool
to
maybe
have
as
like
a
little
lightning
talk
on.
Actually
one
of
our
kick
lightning
talk
webinars
if
as
long
as
we're
not
teasing
them-
and
they
actually
could
use
it
like.
B
C
B
So
yeah,
so
we
decided
we
we
looked
and
we
thought
that
this
should
be
like
a
practical
like
there
should
be
something
that
comes
out
of
this,
that
somebody
could
actually
use.
So
what
we
did
is
we
have
built
a
portal
that
allows
you
to
go
and
search
all
of
this
data
using
a
common
model
so
that
you
can
do
data
discovery
across
all
of
these
different
data
sets.
So
first,
I'm
going
to
give
you
an
extremely
short
tour
of
the
model.
B
Just
so
you
can
kind
of
see
what
we
did
and
then
I'm
going
to
show
you
a
really
short
demo
of
the
portal
and
give
you
the
address
so
that
you
can
use
it
if
you
want
to,
and
then
I'm
going
to
stop
talking
so
whirlwind
tour
of
the
model.
So
what
we
did
is
year,
one
of
this
project.
We
went
to
each
of
those
programs
that
I
showed
you
the
logo
for
we
went
and
physically
visited
them.
B
We
talked
to
them
for
a
couple
of
days,
asked
them
everything
we
could
think
of
about
their
data
and
how
they
store
it
and
what
they
do
to
try
to
figure
out
what
kind
of
model
would
work,
and
then
we
built
a
model
that
everyone
can
use
to
describe
their
data
at
their
portal
or
at
their
place.
So
we
don't
move
the
data
we
host
a
metadata
search
system
and
right
now
we're
working
on
making
it
go
and
retrieve
those
data
files
and
put
them
somewhere
for
you.
C
B
Working
on
inter-operating,
we
are
working
on
being
able
to
send
them
to
workspaces
like
kavatika
to
work
on,
but
in
our
portal
you
can
save
your
search
and
all
of
the
information
about
it
to
your
personal
profile,
very
cool.
Thank
you
yeah.
So,
right
now
we're
in
the
three
quarters,
or
so
through
year,
two,
where
we're
building
this
metadata
model
thing.
B
B
So
if
you
are
interested
in
looking
for
biosamples
that
are
available
through
the
common
fund,
then
we
basically
have
a
file
that
the
person
that
has
all
of
this
data
fills
in
and
it
makes
it
searchable,
and
the
important
thing
about
this
is
that
we
are
trying
to
not
make
any
new
standards.
So
our
model
is
really
sort
of
a
monsterization
of
a
whole
bunch
of
standards
that
already
existed,
we're
just
jamming
them
together
into
something
usable.
B
B
We
use
persistent
ids
and
those
can
be
whatever
persistent
ids
they
already
use,
so
it
can
be
dois
min
ids,
whatever
whatever
they
already
have.
We
support
all
of
them
and
then,
when
we're
describing
the
actual
samples,
we
are
using
controlled
vocabularies
that
already
exist
so,
for
instance,
to
describe
those
different
tissues.
B
We
are
using
the
uber
on
controlled
vocabulary,
which
has
a
gazillion
anatomy
terms,
and
so
we
just
pull
in
uber
on
directly,
and
everyone
takes
whatever
model
they
normally
use
to
describe
their
data
maps
it
as
the
expert
on
their
data
to
uber
on
that
goes
into
our
portal,
and
that
gives
end
users.
The
ability
to
search
across
all
of
these
data
sets
using
the
uber
on
terminology,
even
if
it's
not,
even
though
that
isn't
what
all
of
the
common
fun
data
sets
used
originally.
B
B
I
said
all
of
this
without
clicking
sorry
about
that,
and
we
did
the
same
thing
for
all
of
these
different
things
so,
like
I
said
you
can
also
search
by
file
same
deal,
local
ids
names.
We
can
let
you
search
by
format
by
the
data
type
by
assay
type
by
mime
type
and
the
same
thing
for
each
one
of
these.
We
are
using
existing
ontologies
and
not
making
up
anything
on
our
own.
B
So
you
can
do
sort
of
high
level
what
kind
of
data
exists
if
you
want
to
go
and
find
information,
if
you
want
to
go
and
actually
find
those
files,
you
click
on
the
thing
that
you
would
like
to
locate
and
we
have
a
faceted
search
system
that
lets
you
search
on
any
of
those
different
things.
So,
if
you're
looking
for
say
expression
data
and
you
want
it
from
the
nervous
system,
you
can
find
out
what
kind
of
different
files
we
have.
B
C
B
B
B
That
also
probably
is
just
an
rna
sequence
or
a
dna
sequence
or
a
protein
sequence.
And
so
what
sort
of
happened
is
that,
even
though
we
gave
people
specific
ontologies
to
work
with
they
all
sort
of
picked
and
choose
within
those
ontologies
in
slightly
different
ways,
so
what
we
ended
up
with
was
that
each
program
actually
has
almost
entirely
non-overlapping
data.
So
this
is
all
from
one
program.
This
is
all
from
another
program
and
this
sort
of
flies
in
the
face
of
that
plot.
B
B
B
But
we
have
added
a
second
layer
of
information
that
you
can
search
by.
That
is
perhaps
more
useful
and
that
sort
of
brings
back
that
harmonization
thing
that
we
wanted
to
do
so
now.
If
I
look
for
things
like
connective
tissue
in
the
anatomy
slim,
it
will
bring
me
back
all
of
the
things
that
are
under
that
umbrella
term,
so
that
users
can
see
what
all
of
the
connective
tissues
were,
regardless
of
what
the
specific
term
was
that
the
program
chose-
and
this
is.
C
D
Yes,
I
thank
you,
florence
and
amanda.
It's
great
work,
I'm
not
in
this
area
at
all,
but
just
very
interesting
hearing
about
it.
I
have
some
questions,
not
sure
I
asked
the
right
way
sure
so
from
the
get-go.
The
idea
of
kind
of
putting
this
data
together,
making
it
sustainable,
is
great
and
all
this
multi
kind
of
very
great
green
scale
kind
of
projects.
What,
if
you
have
a
small
data,
provide
a
small
project
right
and
they
may
not
have
abundant
resources.
D
They
want
to
contribute
with
this,
but
because
they
don't
have
that
experience
in
large
scale,
they
may
not
code
their
data.
The
same
way,
the
right
way.
How?
How
would
you
deal
with
those
kind
of
cases
I'm
asking,
because
I'm
thinking
about
more
for
cyber
security
kind
of
data,
but
but
I'm
curious
how
you
deal
with
that
kind
of
situation.
A
Yeah,
so
sorry,
I'm
going
to
moderate
our
privilege
just
for
a
second
I'm
going
to
jump
in
and
just
ask
if
amanda
you
can
start
to
wrap
up
and
then
stop
sharing,
so
kenton
can
get
set
up
thanks
because
we're
we're
we're
pushing
time.
No,
no!
This
is
fantastic.
Thank
you.
I
I
so
go
ahead
and
I'll.
Please
please
go
ahead
and
address
jay's
question
and
we'll
and
we'll
see
if
we
can
get
kenton
set
up
here.
B
B
But
you
can't
just
expect
that
if
you
take
data
and
have
people,
if
you
take
a
model
and
say
here,
use
this
model
use
these
ontologies
that
it
will
automatically
make
your
data
fair.
You
really
have
to
work
with
the
humans
that
are
making
those
decisions,
because
they're
all
going
to
put
their
own
layer
of
humanness
on
how
they
work
with
those
ontologies.
D
B
A
A
There
are
actually
other
technologies
being
used
as
well
that
engage
the
community
to
develop,
what's
called
a
fair
implementation
profile,
so
they
get
together
and
actually
across
can
identify
the
standards
they're
going
to
use
to
help
do
this,
some
of
that
interoperability,
so,
but
all
that
for
another
day,
so
let
me
introduce
kenton.
He
is
associate
director
for
software
at
ncsa.
The
national
center
for
supercomputing
applications
at
uiuc
for
the
last
12
years
he's
helped
to
establish
a
team
of
research
software
engineers
to
support
the
scientific
community.
A
E
You
melissa,
I'm
assuming
everybody
can
see
my
screen,
yes,
okay,
good.
So
what
I'm
going
to
talk
about
today
is
the
earthquakes
effort
and
a
little
bit
about
that
program.
To
begin
with,
so
I
believe
many
of
you
probably
know
the
earth
cube
program
out
of
nsf
was
started
around
2012,
which
is
a
good
decade
ago
now,
and
it
was
started
to
basically
try
to
advance
the
cyber
infrastructure
and
practices
within
geosciences.
E
This
has
basically
been
focusing
around
the
idea
of
developing
cyber
infrastructure
data
resources
tools,
standards
by
which,
together
to
bring
together
and
use
data
better
within
the
geoscience
domain.
So
over
those
that
decade,
roughly
95
efforts
have
been
funded
spanning
all
these
different
areas,
developing
tools
and
resources.
E
So
with
with
anything
like
this,
you
know
there's
a
common
set
of
challenges
that
always
seem
to
arise.
The
most
common
is
around
sustainability,
so
we
fund
these
things
and
then
you
know
what
happens
after
the
funding
ends.
What
happens
to
the
if,
in
the
case
of
a
tool,
what
happens
to
that
tool?
Can
you
still
use
it
exposure?
You
know
being
knowledgeable?
E
What's
out
there
what's
been
funded
and
what's
available
for
me
as
a
researcher
to
actually
leverage,
that's
always
been
a
challenge
in
the
nsf
world
and
what
that
leads
to
is
a
lot
of
duplication
of
effort.
Things
get
built
over
and
over
again,
and
so
things
like
that
tracking,
relevant
products
and
even
packaging
things
so
that
they
are
more
readily
reusable.
E
That's
that's
tricky
at
times
you
know
having
to
install
complex
environments
and
things
like
that
in
order
to
use
things
having
the
necessary
resources
stuff
like
that,
so
these
are
pretty
common
challenges.
With
regards
to
any
of
these
efforts
and
earthquake
has
definitely
seen
these
with
the
with
the
projects.
That's
funded,
you
know,
95
different
efforts,
you
know
what
happens
after
each
of
those
ends.
E
The
idea
is
to
keep
using
these
things,
so
I'm
going
to
talk
a
little
bit
about
the
activities
that
have
been
done
around
that
space,
and
this
includes
both
data
and
tools,
and
so
first,
I
would
just
want
to
kind
of
highlight
the
tech
team.
That's
part
of
the
current
office,
the
eco
office
for
earth
cube.
We
basically
support
the
various
activities
out
there.
E
Try
to
coordinate
activities
with
regards
to
data
and
in
tools
and
foster
communication
and
collaboration,
and
the
annual
meeting
that
happens
each
year
and
it's
one
of
the
kind
of
highlight
that
these
folks
come
from
four
different
institutions
with
a
vast
amount
of
experience
in
the
nsf
world
of
cyber
infrastructure,
having
developed
a
number
of
things
within
dibs
and
cssi
and
so
forth
as
well.
So
there's
a
lot
of
experience
at
hand
with
this.
E
So
on
the
data
side,
the
activity
that
is
called
geo
codes,
and
so
what
it
basically
is
is
this
idea.
This
came
from
work
with
within
our
cube
of
the
council
of
data
facilities.
This
is
the
body
of
repositories
that
hold
some
of
the
larger
data
sets
in
geoscience,
coming
together
with
the
the
drive
to
basically
make
their
data
more
findable
and
usable,
and
so
what
came
out
of
that
was
this
idea
of
adopting
schema.org,
I'm
sorry
yeah
skeet
schema.org,
and
what
that
basically
is
is.
E
It
means
it
was
developed
by
google
and
a
number
of
other
companies
to
basically
annotate
web
pages
landing
pages
with
regards
to
metadata.
That's
present
there,
and
so
the
idea
within
geoscience,
and
specifically
the
science
on
schema
effort,
is
to
leverage
that
to
annotate
data
sets,
and
the
point
of
that
is
so
that
they're
crawlable.
E
That
kind
of
like
google
does
with
web
pages
something
to
come
along
and
discover
this
data
set
and
this
metadata
on
each
page
and
build
a
kind
of
a
master
index
so
that
somebody
could
come
through
and
basically
search
across
a
union
of
different
data
sets.
And
so
the
idea
was
to
leverage
this.
It's
convenient
in
the
sense
that
it
leverages
common
web
technologies
such
as
json
ld.
E
E
So
this
is
the
the
interface
that
was
developed
and
what
I'll
do
is
I'll,
just
kind
of
walk
through
a
few
examples
of
this.
You
can
go
to
it
yourself
here.
If
you
go
to
geocodes.earthcube.org,
but
then
you
could
do
a
few
examples
of
different
things.
So
one
you
could
you
know
a
researcher
comes
by
and
is
looking
for
high
resolution
lidar
data
for
the
july
2019
risk
earthquake.
E
E
All
right-
and
so
what
we
have
here
on
the
left
here
is
a
number
of
facets.
It's
a
number
of
them
collapsed
at
first
to
kind
of
not
bombard
the
user
with
the
ways
they
can
navigate
the
data,
but
they
are
here
if
they
want
to
get
into
more
depth,
so
you
can
do
searches
based
on
resource
types.
So
there's
both
data
and
tools
and
I'll
quickly
mention
that
the
tools
are
coming
from
a
number
of
different
sources.
E
One
of
the
key
efforts
out
of
earthquake
over
the
years
was
the
resource
registry.
So
that's
the
main
source
of
where
these
tools
are
coming
from,
but
then
you
can
search
on
keywords
locations,
the
the
specific
repository
it's
coming
from
so,
for
example,
like
hydroshare
or
others
down
here
and
so
forth,
and
on
the
right.
You
have
your
your
results
and
you
can
see
in
the
bottom
of
the
results,
there's
a
little
bullets
that
indicate
whether
it's
data
and
some
of
them.
E
If
you
scroll
down
are
tools
and
whether
or
not
they
have
a
kind
of
a
connected
tool
with
it,
and
this
is
the
kind
of
the
key
aspect
of
the
tools
so
that
you
can
kind
of
find
the
tools
that
are
used
with
that
data
set.
So
if
you
click
on
one
of
the
results
you
can
get,
you
know
the
usual
metadata,
an
abstract
about
it's
a
doi.
E
E
If
it's
the
carrying
the
tools
where
you
have
to
basically
download
sorry,
I'm
trying
to
move
my
my
bar
here,
so
I
can
get
back
to
my
presentation
and
and
if
you
scroll
down
even
further
you'll,
see
related
data
sets,
and
this
again
is
based
on
metadata
kind
of
overlaps.
There
you
can
click
on
those
get
some
summaries
of
the
event
data,
and
if
you
actually
want
to
go
to
the
data
set,
then
you
can
actually
click
on
that
and
it'll.
E
E
In
this
case,
what
I
want
to
highlight
is
again
in
the
kind
of
connected
tools
when
they
do
pop
up.
Some
of
them
are
web
tools,
so
in
this
case
there's
one
of
those
it
shows
here
as
a
web
application
and
if
you
click
on
that,
it'll
actually
take
you
to
the
web
tool
and,
if
possible,
populate
it
with
the
data
that
you
were
just
looking
at
so
this
case
it
has
that
information.
E
So
if
you
want
to
try,
it
feel
free,
you
can
just
go
here
and
try
it
out.
This
is
the
current
version.
We
keep
it
up
to
date,
as
we
add
new
features
to
it
on
the
tools
side,
we
are
looking
at
other
ways
of
basically
getting
tools.
One
of
the
ways
we're
looking
at
is
basically
leveraging
one
of
the
earthquakes
that
was
funded
called
throughput.
E
Other
things
we've
been
working
on
in
terms
of
capabilities
is
the
ability
to
kind
of
scroll
through
time
so
adding
like
a
scroll
bar
here
with
regards
to
time
in
this
case
publication
year,
but
also
potentially
the
time
with
the
ghost
of
the
data
set
and
additionally,
the
addition
of
geospatial
capabilities
of
searching.
So
maybe
a
map
where
you
kind
of
draw
boxes
and
find
data
sets
within
that
specific
area
which
would
be
useful
for
geoscience
research
and
then
the
addition
of
notebooks.
E
I'm
gonna
skip
this
for
lack
of
time,
but
with
regards
to
the
actual
tools,
and
so
one
of
the
things
we've
challenges
we've
had
here
also
is
with
regards
to
actually
identifying
those
tools
and
making
them
more
useful,
useful
and
so
again,
with
regards
to
the
data
set
search.
You
know
that's
a
useful
thing
too,
in
terms
of
being
able
to
use
data,
and
so
over
the
years
with
regards
to
tools,
there
have
been
a
number
of
fronts
in
regards
to
how
to
do
that.
E
It
was
also
this
idea
of
a
workbench
that
was
being
thrown
around
by
the
previous
office
and
what
that
was
is
basically
a
kind
of
different
kind
of
resource
registry,
where
tools
were
also
usable
right
from
the
web
page
and
maybe
connectable
in
terms
of
making
workflows
or
tools
and
so
forth,
and
so
that
was
one
thing
that
was
being
bounced
around
as
a
direction
to
possibly
going,
but
there's
complications
with
regards
to
that.
There's
also,
this
big
push
for
notebooks
that's
happening
throughout
science,
and
so
the
idea
too
was
you
know.
E
Could
we
do
something
there
to
kind
of
really
leverage
notebooks
jupiter,
our
markdown,
you
name
it,
and
so
in
addition
to
that,
there's
also
this
growing
push
for
software
as
a
scholarly
object.
So
it
is
in
itself.
You
know
scientific
output,
it's
something
worth
citing,
and
so
there's
a
growing
movement
around
that
and
then
from
nsf's
side.
There's
always
this
push
for
sustainability.
E
We
need
to
keep
these
activities
whatever
they
are
going
long
term
and
there's
this
drawback
with
regards
to
this
central
registry
kind
of
aspect
is
that
somebody
needs
to
take
care
of
it,
and
so
who
is
that?
How
do
you
fund
that
and
how
do
you
keep
it
going,
and
so
one
of
the
interesting
notions
that
came
out
from
a
colleague
of
mine,
clara
narc,
said
at
the
university
of
illinois
was
with
regards
to
getting
a
community
to
do
anything.
E
It's
important
to
really
incentivize
the
community
in
that
regards,
and
in
regards
to
something
like
this,
the
aspect
of
something
that
could
be
on
their
cv,
something
worth
citing
and
what
really
came
out
of
that
was
this
idea
of
well.
Could
we
capture
tools
as
notebooks,
if
they're
peer-reviewed,
and
by
doing
the
peer-review
aspect,
to
incentivize
the
community
to
do
it
for
us
versus
us
doing
it
for
them?
We,
you
know
if
we
had
to
build
all
the
notebooks
for
all
the
tools
out
there
kind
of
documenting
them
and
capturing
how
they
run.
E
We
didn't
really
know
if
this
would
work
or
not
or
if
anybody
would
be
interested
at
all,
and
so
me,
dave,
volker
lynn,
schreiber
and
dan
katz
who's.
One
of
the
founders
of
josh,
the
journal
of
open
source
software
worked
on
this
and
issued
this
call.
There
was
pleasantly
some
interest
from
the
community.
There
was
some
twitter
activity
around
it.
E
E
During
the
review
process,
a
number
of
viewers
had
trouble
running
some
of
them,
so
we
wanted
to
address
that
if
we
ever
did
this
again-
and
there
was
this
clear
focus
on
science
aspects
of
of
stuff
versus
the
tools
aspect
of
stuff,
some
were
showcasing
tools.
Some
were
showcasing
the
science
questions
that
just
happened
to
use
a
tool,
so
you'll
see
different
science
questions
being
answered
with
the
same
tool,
so
we
felt
the
need
to
kind
of
address
that
and
then
formatting
we
did
with
the
conference.
E
So
we
we
tried
that
and
then
we
ended
up
having
two
tracks
too
one
for
research,
education
and
one
for
demonstration
and
tools
as
well.
So
we
did
that
and
expanded
the
organizing
committee
and
the
reviewers,
because
we
ended
up
having
way
more
submissions,
so
we
ended
up
having
44
abstracts
submitted,
which
is
kind
of
like
an
intent
to
submit
a
notebook.
We
only
had
33
final
submissions
was
still
more
than
last
year.
E
We
think
that
maybe
there
wasn't
enough
time
between
the
abstract
submissions
and
the
final
submissions
and
so
some
dropped
out,
but
we
had
a
lot
more
from
students
and
early
career
folks,
which
is
a
good
thing,
is
what
you
would
expect
with
any
kind
of
conference
kind
of
proceedings,
and
we
ended
up
accepting
19
this
time,
11
as
oral
presentations,
and
so
the
experiment
kind
of
really
worked.
Well,
we
ended
up
publishing
it
this
time
with
a
jupiter
book,
which
is
a
new
technology
that
jupiter
is
pushing
out.
E
It
kind
of
makes
a
group
of
jupiter
notebooks
look
like
a
proceedings
which
is
convenient
for
us
and
then
using
github
and
zenodo
to
kind
of
store
the
notebooks
and
get
minting
them
doi's
one
another
cal,
poly
insights
came
out
of
this
as
well,
though
one
is
an
emphasis
on
testing.
We
had
quite
a
few
notebooks
that
were
that
worked
when
they
were
submitted,
but
didn't
work
when
they
were
being
reviewed,
and
we
worked
with
the
submitters
to
kind
of
address
those
if
we
could,
but
it
was
it
raised
some
issues.
E
You
know
that
are
common
with
software
development.
It's
the
importance
of
testing
and
freezing
things.
So
that's
something
interesting.
If
publications
go
down
that
route
of
software,
but
then
again
also
we
we
got
some
feedback
from
the
community
that
they
would
like
not
just
to
be
restricted
to
jupiter,
which
was
interesting
because
the
previous
year
was
all
jupiter,
but
so
the
idea
to
support
again
our
markdown
as
a
possibility,
even
matlab,
for
submissions
in
the
future.
E
Is
those
tools
go
into
basically
our
resource
registry?
It
helps
us
populate
in
more
tools
that
then
get
associated
with
data
sets
in
geo
codes,
and
so
we're
really
trying
to
push
that
that
angle
of
tools
of
data
kind
of
going
together
and
really
coming
from
these
community
resources.
Data
repositories
called
by
schema.org
or
notebooks
that
are
provided
by
the
community
and
so
what's
come
out
of.
E
This,
too,
is
a
kind
of
a
best
practices
document
put
together
by
the
earth,
cube
leadership,
council
tech
in
office
that
kind
of
outlines
these
practices
over
these
principles.
So
I
guess
over
any
specific
way
of
doing
things.
Project
wise
is
basically
adoption
of
standards
like
schema.org
adoption
of
notebooks
and
binder,
and
things
like
that,
just
to
make
things
more
usable
and
more
fair,
so
I'll
close
there
take
any
questions
done.
My
thing.
A
So
I
just
have
one
quick
question
and
then
we'll
wrap
up
for
today.
I
know
there's
a
couple
of
announcements
and
I
need
to
check
something
with
with
with
chris
daley
before
we
close
so
kenton.
I
know
that
you've
been
working
with
github
and
and
zenodo
to
make
sure
that
there's
a
collection
of
these
that
will
be
long.
You
know
in
a
regular
repository.
A
E
I
don't
see
it
as
having
to
be
one
or
the
other.
We
are
actively
working
with
agu
and
they
are
actively
working
on
supporting
this
idea
of
notebooks
or
scholarly
objects,
so
they
are
definitely
going
forward
with
that
direction
as
well.
But
I
I
see
you
know
leveraging
things
like
github
and
zenoto,
and
jupiter
book
is
another
way,
especially
for
smaller
organizations.
To
do
this
also
so.
D
Going
back
to
jay's
question
ken,
I'm
wondering
if
you
could
just
say
a
little
bit
about
how
the
tech
team
has
worked
with
the
different
data
facilities,
large
and
small,
to
try
and
help
them.
You
know,
implement
some
standards
and
because
that
might
be
an
interesting
model
for
the
stuff
that
he's
working
on.
E
Yeah
that
rang
a
bell
from
the
previous
presentation
is
that's
a
big
part
of
what
the
tech
team
in
the
eco
office
does,
and
I
imagine
the
previous
esso
office
as
well
is
with
each
of
the
repositories
in
the
cdf.
There
has
been
a
lot
of
interaction
with
regards
to
you
know
supporting
the
schema.org,
but
also
data
cleaning
making
it
so
that
things
just
kind
of
work
together,
and
so
that's
an
important
part
of
this
kind
of
activity.