►
From YouTube: 2019-07-09 : DSpace 7 Entities Working Group Meeting
B
B
B
Regarding
the
this
desired,
the
pay
our
priorities
in
our
list,
so
we
we
just
created
and
I-
will
try
to
get
back
to
this
one.
We
created
a
mock-up
for
the
submission
process.
How
can
we
integrate
entities
in
the
submission
process?
Well,
how
can
we
relate
person
with
a
publication?
This
is
the
the
use
case
that
we
study
and.
B
B
A
If
we
look
something
very
high-level,
we
have
a
relation
between
a
publication
on
the
outer
and
then,
if
we
look
further,
we
need
to
specify.
First
is
one
of
the
use
cases
regarding
the
outer
name
variants.
So
someone
has
a
specific
name
for
a
particular
publication,
but
also
it
may
have
a
specific
affiliation
on
that
particular
publication.
A
So
the
question
you
hear
is
through
all,
maybe
use
the
same
approach
for
these
two
situations,
because
in
both
we
will
have
a
relation
between
the
person
and
the
publication
that
has
some
kind
of
intermediate
attributes
between
the
publication
on
the
outer.
So
it's
it's
an
auto
related
to
publication
with
some
specific
attributes.
That
is
a
specific
name
and
specific
affiliations.
A
D
B
C
So
the
link
between
FCT
and
the
publication
is
a
direct
link
between
those
two
entities.
So
it
doesn't
go
through
the
person,
so
the
link
will
never
change
and
the
only
thing
that
you
don't
know
in
that
case
is
which
author
has,
which
affiliation,
because
here
you
have
two
affiliations:
FCT
and
Universidad
Dominion
and
you
wouldn't
know
which
of
the
author's
belongs
to
which
organizational
unit
yeah.
Sorry.
C
This
case,
you
would
have
a
link
between
drama
and
thus
more
era
and
the
publication,
and
you
would
also
have
a
link
between
FCT
and
publication.
The
only
thing
that
wouldn't
know
is
that
Joelle
is
from
FCT,
where
Jose
Carvalho
is
from
from
minion,
so
that
that
would
be
the
only
thing
that
you
don't
know.
But
the
question
is:
do
you
need
to
know
that
for
any
particular
report.
B
C
You
yeah
so,
but
you
could
do
a
query
that
says,
give
me
all
the
publications
related
to
Katya
and
only
consider
the
ones
that
have
an
affiliation
tab.
City,
France,
no
okay,
but
you
don't
know
for
sure
whether
the
link
between
AB
see
whether
Katya
was
related
to
F
city
or
some
other
organization.
At
that
time,
that's
something
you
do!
Yeah.
B
B
C
You
know
it
very
much
depends
on
which
kind
of
reports
you
need
to
generate.
So
if
you
need
to
know
okay,
how
many
publications
were
linked
to
FCT
and
funded
by
FCT,
you
can
do
this
with
the
current
solution.
Yes,
if
you
really
need
to
know
how
many
projects
or
how
many
publications
were
published
by
katya
in
while
she
was
working
at
FCT,
then
this
this
solution
wouldn't
work
because
you're
not
sure,
if
they're
more
than
one
author
and
more
than
one
affiliation
you're,
not
certain
that
a
particular
author
is
affiliated
to
a
particular
institution.
C
So
with
the
current
solution
that
is
more
simple
in
design
and
and
simpler
in
implementation,
you
can
already
generate
quite
a
lot
of
reports,
but
there
are
some
cases
where
you
would
need
to
have
the
link
between
the
author
and
the
organizational
unit,
but
I
have
currently
not
heard
any
example
of
a
particular
report
where
that
is
needed.
So
as
long
as
that
is
not
needed,
we
don't
need
to
have
these
three-way
or
triple
relations
or
adding
the
affiliation
to
the
relation
between
the
author
and
the
publication,
and
that
is
an
implementation.
C
A
B
A
Think
it's
more
the
way
we
implement,
which
is
different
in
the
way
we
we
manage
the
information
we
need
because
we
have
ear
relation
between
and
if
we
consider
a
basic
approach,
we'll
have
a
relation
between
the
publication
and
the
person,
a
person
that
has
some
affiliations
and
then
we'll
have
a
relation
between
a
specific
affiliation
and
the
publication.
What
we
we
need
to
make
sure
to
this
to
work
is
to
have
the
same
information
on
the
relation
between
the
publication
and
the
affiliation
where
the
affiliation
is
the
same
than
the
affiliation
of
the
outer.
A
C
A
So
the
question
is
that
we
have
one
outer-
that
is
part
of
different
departments
and
research
centers.
So
we
take
the
person
John
Doe,
who
has
three
different
affiliations,
a
a
B
and
C
for
a
specific
publication.
He
may
use
one
to
the
tree
or
just
another
one.
Okay,
so
he
may,
depending
on
the
topic,
is
writing
shows
different
definitions.
A
A
So-
and
this
is
the
way
we
we
want
or
needs
to
be
able
to
distinguish
on
that-
that
context
of
managing
the
information
and
repositories
and
and
this
same
context,
for
example
as
alternative
names.
It's
a
concept
that
is
also
usual,
for
example,
for
organizations
where
we
can
have
different
names
associated
with
different
publications.
But
it
connects
to
one
unique
organization
with
a
specific
identifier.
A
C
B
C
Say
what
one
thing
about
this
so
I
understand
that
the
the
workaround
you've
just
presented,
but
the
the
question
still
remains
like
I
would
like
to
see
a
use
case
of
a
report
that
really
requires
this.
That
really
requires
you
to
know
which
person
was
affiliated
to
which
organization
for
a
particular
publication.
So
far,
I
haven't
seen
any
example
report
that
really
requires
okay.
C
E
B
Showing
showing
us
I
will
I
I
would
I
think
we
can
revisit
the
the
the
mock-ups
for
the
submission
process
for
for
those
that
didn't
watch
the
first
time.
This
is
what
we
thought
initially
this.
This
isn't
this
isn't
closed
version,
but
but
is
open
to
discussion
contributions,
and
what
we
thought
in
this
submission
form
was
that
in
the
author's
field
we
could
type
some
name
and
some
results
were
suggested
and
if
none
of
the
the
results
were
the
one
that
I
was
searching
for
I
could
search
for
more
results.
B
B
But
the
idea
was
that
the
user
could
pick
one
of
the
one
of
the
name
variants
and
or
type
one
if
none
of
the
the
author's
name
variant
was
listed
here
or
person,
name
variants
was
listed
here
and
one
other
aspect
that
we
had
I
think
we
can't
better
improve
the
experience
for
the
user
is
to
have
a
button
or
something
that
is
triggers.
The
creation
of
a
new
entity
right
on
this
form
or
such
research
component.
B
B
C
So
we
had
also
been
thinking
about
how
to
design
this,
and
we
were
thinking
more
in
a
kind
of
a
two-step
approach
where,
if
you
click
on
the
author,
John
dill
and
the
system
sees
that
not
click
like
that.
Sorry,
we
selected
well
when
you
would
select
that
author,
that
you
would
then
have
a
second
step
where
the
system
asks
you
in
the
case
of
if
that
author
has
named.
C
Okay,
just
select
the
name
variant,
so
the
main
reason
is
technical,
because
we
would
like
to
be
able
to
reuse
the
way
that
this
rich
component
exists
today
in
as
many
places
as
possible
and
with
what
you
have
designed
here,
you
would
need
to
basically
create
a
new
view
of
the
search
results
which
would
be
only
specific
to
only
this
part
of
the
repository
where,
if
you
just
put
it
in
two
steps,
you
also
let
the
user
think
less.
You
should
ask
for
some
input
on
the
from
the
expert
on
that
as
well.
C
But
that
way
you
don't
have
to
think
so
much
about
the
first
step.
It's
just
like
I
adds
this
John
Doe,
you
click
on
it,
and
then
you
get
the
question
of
which
name
variant.
Would
you
like
to
use,
or
would
you
like
to
add
variant
that
would
also
make
implementation
of
adding
a
new
name
variant
simpler
or
cleaner
than
if
you
would
do
it
all
in
one
list,
but
that
would
be
our
that's
the
direction
we
are
going
in
at
the
moment.
C
And
so
you
can
add
more
authors
more
easily.
But
yeah.
We
that's
a
discussion.
We've
only
had
this
week
and
somebody's
looking
into
our
dua
expert
is
looking
into
the
design
and
we're
also
looking
at
all
of
the
other
use
cases.
For
example,
if
you're
submitting
a
new
person-
and
you
want
to
attach
the
20
publications
that
that
person
already
has
to
to
the
as
relations
to
the
person
and
that
person
could
have
published
it
under
different
names.
C
So
it
could
be
doe,
coma,
John
or
DoCoMo
GA
or
whatever,
so
that
you
can,
just
you
know,
keep
doing
your
searches
at
first
do
a
search
for
Doe,
comma
John
and
then
add
five
publications
to
the
list
and
then
just
do
another
search
and
just
keep
doing
that
while
that
one
window
is
open,
but
we
will
have
a
proposal
as
well,
probably
by
the
next
meeting.
Okay.
B
B
C
B
B
B
E
That
can
definitely
be
implicit,
so
there's
no
need
for
that
to
be
present,
but
the
changing
the
type
attribute
estate
that
it's
a
relationship
would
imply
that
we
no
longer
know
what
the
type
for
the
plaintext
lookups
sort
of
plain
text
entry
would
be.
So
if
you-
and
if
you're,
submitting
all
the
circle
location,
you
would
want
to
know
that
when
you're
not
doing
search
for
an
entity,
you
want
to
be
able
to
see
a
last
name
in
the
first
name
box,
and
that
has
a
type
name
at
the
moment.
E
B
I,
wasn't
the
only
one
III,
don't
think
Alexander,
also
it
to
to
to
to
add
that
block
that
you
just
mentioned
because
of
my
experience
with
other
platforms
and
and
how
they
they
described.
Fields
and
I
thought
that
it
was
better
to
have
a
different
kind
of
field
here,
because
we
all
we
already
have
one
I-
think
it's
two
to
field
or
two
to
text.
I,
don't
I!
Don't
quite
remember
that
I.
B
E
Way
that
the
rest
contract
was
actually
here
approached
here
was
quite
similar
to
a
metal,
a
the
fields
with
not
thority
lookups.
So
if,
if
you'd
check
out
how,
for
instance,
the
author's
are
being
defined
right
now
and
I
included
the
link
to
the
to
the
rest
contract
in
the
chat
here
that
just
states,
okay,
there's
an
input
with
typed
name
and
there's
a
label
for
its
it's
just
a
little
bit
lower.
E
Yes,
so
that
contains
all
of
the
details
and
you
could
have
an
authority
control
lookup
in
there
as
well
and
I,
try
to
keep
it
as
closely
to
this
format
as
possible
and
just
say:
okay,
apart
from
being
able
to
store
metadata
and
visa
contributor
order,
we
can
also
look
up
metadata
from
a
specific
relationship
being
in
this
case.
There
is
order
of
publication
relationship.
B
E
Are
all
well,
each
metadata
field
has
one
input
type
in
name
box,
one
box
text
area
so
that
that's
always
defined,
and
if
you're
mixing
entities,
new
clips
and
methods
and
plain
text
metadata
entry,
it's
still
need
to
know
in
what
kind
of
type
what
type
of
box
you
want
to
enter
your
plain
text
metadata.
Do
you
want
to
use
one
box?
Do
you
want
to
use
a
name
box?
Do
you
want
text
areas
are
probably
not
gonna
be
dead,
like
yeah.
B
D
E
Be
able
to
choose
them
and
keeping
the
structure
the
same
would
make
that
a
lot
easier
to
turn
to
runner
in
the
angular
UI
so
that
you
can
say
okay.
This
is
a
name
box,
but
what's
the
difference
about
this,
there's
the
one?
That's
that's
here
on
the
top
and
the
one
that's
in
the
pull
request.
Only
that
there
is
no
selectable
relationship
where
you
can
say:
okay,
so
want
to
be
able
to
build.
E
B
E
E
E
Tears
the
reusability,
I
think
and
angular
UI,
because
now
you
could
have
a
component
where
you
say:
okay,
I
have
this
name
input
type,
and
it's
rounded
like
this
or
I,
have
nine
books
and
put
typing
it's
relevant
like
this,
and
then
you
have
optionally,
also
a
lookup
feature
to
be
able
to
to
enter
meant
to
enter
there
or
to
search
for
the
related
entities.
I.
B
C
Think
an
additional
problem
with
that
approach
would
be
that
if
you
also
have
a
field
that
could
have
and
authority
control
and
relationships
that
you
need
another
field
specific
or
field
type
specific
for
each
of
those
combinations.
So
you
would
add
up
with
a
lot
of
different
field
types
which
makes
it
also
more
difficult
for
somebody
to
configure
the
system.
Or
now
you
can
just
think
like.
Okay,
I
want
a
name.
If
that's
plain
text,
it
needs
to
happen
the
name,
input
type
and
then
do
I
want
to
add
a
relationship.
Yes
or
no.
C
B
E
E
E
B
B
B
C
C
Here
we
didn't
implement
the
autosuggest
as
you
have
them
as
you
have
it
in
your
mock-ups,
yet
Paulo,
because
it's
just
the
first
version
and
if
you
click
on
the
lookup
button,
you
get
this
modal
window
where
right
now
it's
using
radio
buttons
here,
because
there
is
only
one
issue
that
you
can
select
in
this
case
and
you
see
the
different
issues.
You
have
a
search
query.
C
And
when
you
click
on
the
add
button,
you
actually
add
it
as
a
metadata
value
to
the
submission.
That's
not
kind
of
like
two-step
process,
but
where
we
do
want
to
change
the
design
a
little
bit
more
so
that
the
two-step
process
happens
with
less
clicks
and
not
having
to
close
the
window
and
open
it
again,
etc.
But.
C
Back
to
that
next
week,
this
is
the
alternative,
where
you
have
the
option
of
adding
multiple
issues
to
the
volume.
So
you
have
checkboxes,
you
can
click
on
two
of
them
and
then
this
control
here
behaves
similarly
to
what
you
have
in
Gmail,
for
example,
and
it
also
has
the
option
to
select
everything
on
the
page,
to
deselect
everything
on
this
particular
page
to
select
all
or
to
deselect
all
the
items.
C
So
if
you
click
select
all
that's
what
you
get
and
then
they're
added
as
a
list
here
to
the
to
this
field
and
when
you
click
on,
add
they're
added
as
individual
issues
to
this
particular
volume.
So
just
keep
it
like
this
will.
This
was
just
the
first
step
to
start
working
on
it.
It
will
elaborate
on
this
and
and
get
it
more
in
line
with
what
we
have
in
the
markups
in
the
next
few
steps.
B
B
B
We
can
specify
also
some
qualifiers.
We
have.
This
displace
already
have
done
this
for
other
identifiers
like
yes,
SN
or
something
like
that,
and
we
can
create
create
upfront
some
and
foresee
some
some
of
these
identifiers
or
give
the
the
installers
the
the
the
decision
to
which
one
of
the
which
one
to
use
and
the
last
option
I
think
it's
the
most
complicated,
because
we
don't
have
the
the
control
over
it,
which
is
to
rely
on
schema.org.
B
C
C
Yeah,
sorry
and
I
actually
forgot
to
mention
one
thing:
if
I
can
quickly
have
the
screen
back,
okay,
I
think
these
are
good
options
and
should
be
considered,
but
we
should
leave
it
for
when
we
get
some
more
feedback
from
Tim
Tims
request.
So
yeah
I
just
forgot
one
thing:
small
thing:
that's
something
that
Tim
brought
up
its
energy
Raticate,
but
I
don't
know
if
you
can
paste
a
link
to
the
JIRA
ticket.
C
B
C
That's
one
thing
I'm
and
then
the
second
thing
is
that
the
relations
were
currently
backward
backwards
in
the
sense
that
you
want
to
use
this
author
of
publication
in
the
direction
of
the
author
to
the
publication,
so
that
it
appears
on
the
author
metadata
core
input,
as
this
object
that
you're
describing
is
the
author
and
you're
pointing
to
the
publication.
So
I
just
wanted
to
show
you
how
we're
gonna
change
it.
C
That's
what
we're
gonna
there
has
to
be
an
S
here,
because
it's
leftwards
label
and
rightwards
label,
but
so
leftwards
label
means
that
it's
the
arrow
pointing
left
work
and
the
right
words
label
is
the
arrow
pointing
right
word
so
to
the
right
type
of
the
relationship
and
we're
just
gonna
make
us
a
small
change,
but
it's
affected
and
a
lot
of
different
locations.
So
we
just
want
to
make
sure
that
nobody
has
any
objections
to
this
before
we
make
that
change,
because
it
does
involve
touching
a
lot
of
files.
C
C
Okay,
do
you
see
the
second
one?
Now?
Second
tab?
Prison,
that's
weird!
It's
not
changing!
Now
it's
weird!
If
I
change
tabs,
apparently
that
doesn't
work
okay,
this
is
gonna,
be
the
final
version.
Okay,
so
you
have
left
words
label
which
basically
says
that
it's
pointing
to
the
left
type.
So
it's
pointing
leftward
and
the
right
words
label
is
pointing
right
word
to
the
right
type
of
the
relation,
and
so
you
would
also
have
on
the
publication
page.
C
You
would
see
its
publication
of
author
so
that
publication
comes
first,
because
that's
the
object,
you're
describing
and
it's
pointing
to
an
author
and
on
the
author
page
it
would
be
publication,
his
author
of
publication,
meaning
that
you
describe
the
author
and
you're
pointing
to
the
publication.
So
it's
more
clear
that
the
left
words
label
points
to
the
left
type
and
the
right
words
label
points
to
the
right
types
of
I
agree.
B
B
C
It's
more
clear
like
the
before
it
was
left
label
and
right
label
where
we
meant
that
the
left
label
points
to
the
left
object.
But
if
you
say
left
word,
then
it's
like
an
English
native
speaker
is
easier
to
understand,
but
it's
pointing
to
the
left
anyway.
Yeah,
it's
just
a
small
thing,
because
I'm
aware
at
the
time
it's
the
if
there
are
no
objections,
we'll
just
make
that
change
and
make
it
more
clear.
B
D
C
E
C
C
B
B
Okay,
I
will
self
assign
the
the
ones
that
need
a
second
reviewer
and
I
will,
because
we
don't
have
any
more
time.
I
will.
Thank
you
all
for
having
again
the
time
to
discuss
this
and
for
contributing
to
to
this
meeting.
Sorry
for
anything
that
didn't
go
as
well,
so
I
will
see
you
on
I
think
Friday
for
a
Thursday,
some
of
you
and
next
week,
with
all
expect.