<s>
Dataspaces	B-Application
are	O
an	O
abstraction	O
in	O
data	B-General_Concept
management	I-General_Concept
that	O
aim	O
to	O
overcome	O
some	O
of	O
the	O
problems	O
encountered	O
in	O
data	B-General_Concept
integration	I-General_Concept
system	O
.	O
</s>
<s>
The	O
aim	O
is	O
to	O
reduce	O
the	O
effort	O
required	O
to	O
set	O
up	O
a	O
data	B-General_Concept
integration	I-General_Concept
system	O
by	O
relying	O
on	O
existing	O
matching	O
and	O
mapping	O
generation	O
techniques	O
,	O
and	O
to	O
improve	O
the	O
system	O
in	O
"	O
pay-as-you-go	O
"	O
fashion	O
as	O
it	O
is	O
used	O
.	O
</s>
<s>
Labor-intensive	O
aspects	O
of	O
data	B-General_Concept
integration	I-General_Concept
are	O
postponed	O
until	O
they	O
are	O
absolutely	O
needed	O
.	O
</s>
<s>
Traditionally	O
,	O
data	B-General_Concept
integration	I-General_Concept
and	O
data	B-General_Concept
exchange	I-General_Concept
systems	O
have	O
aimed	O
to	O
offer	O
many	O
of	O
the	O
purported	O
services	O
of	O
dataspace	B-Application
systems	O
.	O
</s>
<s>
Dataspaces	B-Application
can	O
be	O
viewed	O
as	O
a	O
next	O
step	O
in	O
the	O
evolution	O
of	O
data	B-General_Concept
integration	I-General_Concept
architectures	O
,	O
but	O
are	O
distinct	O
from	O
current	O
data	B-General_Concept
integration	I-General_Concept
systems	O
in	O
the	O
following	O
way	O
.	O
</s>
<s>
Data	B-General_Concept
integration	I-General_Concept
systems	O
require	O
semantic	O
integration	O
before	O
any	O
services	O
can	O
be	O
provided	O
.	O
</s>
<s>
Hence	O
,	O
although	O
there	O
is	O
not	O
a	O
single	O
schema	O
to	O
which	O
all	O
the	O
data	O
conforms	O
and	O
the	O
data	O
resides	O
in	O
a	O
multitude	O
of	O
host	O
systems	O
,	O
the	O
data	B-General_Concept
integration	I-General_Concept
system	O
knows	O
the	O
precise	O
relationships	O
between	O
the	O
terms	O
used	O
in	O
each	O
schema	O
.	O
</s>
<s>
As	O
a	O
result	O
,	O
significant	O
up-front	O
effort	O
is	O
required	O
in	O
order	O
to	O
set	O
up	O
a	O
data	B-General_Concept
integration	I-General_Concept
system	O
.	O
</s>
<s>
Dataspaces	B-Application
shift	O
the	O
emphasis	O
to	O
a	O
data	O
co-existence	O
approach	O
providing	O
base	O
functionality	O
over	O
all	O
data	O
sources	O
,	O
regardless	O
of	O
how	O
integrated	O
they	O
are	O
.	O
</s>
<s>
For	O
example	O
,	O
a	O
DataSpace	B-Application
Support	O
Platform	O
(	O
DSSP	O
)	O
can	O
provide	O
keyword	B-Application
search	I-Application
over	O
all	O
of	O
its	O
data	O
sources	O
,	O
similar	O
to	O
that	O
provided	O
by	O
existing	O
desktop	O
search	O
systems	O
.	O
</s>
<s>
When	O
more	O
sophisticated	O
operations	O
are	O
required	O
,	O
such	O
as	O
relational-style	O
queries	O
,	O
data	B-Application
mining	I-Application
,	O
or	O
monitoring	O
over	O
certain	O
sources	O
,	O
then	O
additional	O
effort	O
can	O
be	O
applied	O
to	O
more	O
closely	O
integrate	O
those	O
sources	O
in	O
an	O
incremental	O
fashion	O
.	O
</s>
<s>
Similarly	O
,	O
in	O
terms	O
of	O
traditional	O
database	O
guarantees	O
,	O
initially	O
a	O
dataspace	B-Application
system	O
can	O
only	O
provide	O
weaker	O
guarantees	O
of	O
consistency	O
and	O
durability	O
.	O
</s>
<s>
Data	O
graphs	O
play	O
an	O
important	O
role	O
in	O
dataspaces	B-Application
systems	O
.	O
</s>
<s>
Search	O
and	O
relational-style	O
queries	O
and	O
analytics	O
can	O
work	O
simultaneously	O
on	O
data	O
graphs	O
which	O
is	O
another	O
important	O
property	O
of	O
dataspaces	B-Application
.	O
</s>
<s>
The	O
goal	O
of	O
personal	B-Application
information	I-Application
management	I-Application
is	O
to	O
offer	O
easy	O
access	O
and	O
manipulation	O
of	O
all	O
of	O
the	O
information	O
on	O
a	O
person	O
's	O
desktop	O
,	O
with	O
possible	O
extension	O
to	O
mobile	O
devices	O
,	O
personal	O
information	O
on	O
the	O
Web	O
,	O
or	O
even	O
all	O
the	O
information	O
accessed	O
during	O
a	O
person	O
's	O
lifetime	O
.	O
</s>
<s>
Our	O
desktops	O
typically	O
contain	O
some	O
structured	O
data	O
(	O
e.g.	O
,	O
spreadsheets	B-Application
)	O
and	O
there	O
are	O
important	O
associations	O
between	O
disparate	O
items	O
on	O
the	O
desktop	O
.	O
</s>
<s>
Finally	O
,	O
we	O
would	O
like	O
to	O
query	O
about	O
sources	O
,	O
e.g.	O
,	O
"	O
find	O
all	O
the	O
papers	O
where	O
I	O
acknowledged	O
a	O
particular	O
grant	O
,	O
"	O
"	O
find	O
all	O
the	O
experiments	O
run	O
by	O
a	O
particular	O
student	O
,	O
"	O
or	O
"	O
find	O
all	O
spreadsheets	B-Application
that	O
have	O
a	O
variance	O
column.	O
"	O
</s>
<s>
Groups	O
will	O
need	O
to	O
federate	O
with	O
other	O
groups	O
to	O
create	O
scientific	O
dataspaces	B-Application
of	O
regional	O
or	O
national	O
scope	O
.	O
</s>
<s>
Users	O
of	O
the	O
federated	O
dataspace	B-Application
may	O
want	O
to	O
see	O
collections	O
of	O
data	O
that	O
cut	O
across	O
the	O
groups	O
in	O
the	O
federation	O
,	O
such	O
as	O
all	O
observations	O
and	O
data	O
products	O
related	O
to	O
water	O
velocity	O
,	O
or	O
all	O
data	O
related	O
to	O
a	O
certain	O
stretch	O
of	O
coastline	O
for	O
the	O
past	O
two	O
months	O
.	O
</s>
<s>
a	O
dataspace-wide	O
catalog	O
,	O
</s>
