<s>
A	O
distributional	B-General_Concept
–	I-General_Concept
relational	I-General_Concept
database	I-General_Concept
,	O
or	O
word-vector	B-General_Concept
database	O
,	O
is	O
a	O
database	B-Application
management	I-Application
system	I-Application
(	O
DBMS	O
)	O
that	O
uses	O
distributional	O
word-vector	B-General_Concept
representations	O
to	O
enrich	O
the	O
semantics	O
of	O
structured	B-Application
data	I-Application
.	O
</s>
<s>
As	O
distributional	O
word-vectors	O
can	O
be	O
built	O
automatically	O
from	O
large-scale	O
corpora	O
,	O
this	O
enrichment	O
supports	O
the	O
construction	O
of	O
databases	B-Application
which	O
can	O
embed	O
large-scale	O
commonsense	O
background	O
knowledge	O
into	O
their	O
operations	O
.	O
</s>
<s>
Distributional-Relational	O
models	O
can	O
be	O
applied	O
to	O
the	O
construction	O
of	O
schema-agnostic	B-General_Concept
databases	I-General_Concept
(	O
databases	B-Application
in	O
which	O
users	O
can	O
query	O
the	O
data	O
without	O
being	O
aware	O
of	O
its	O
schema	B-Application
)	O
,	O
semantic	O
search	O
,	O
schema-integration	O
and	O
inductive	O
and	O
abductive	O
reasoning	O
as	O
well	O
as	O
different	O
applications	O
in	O
which	O
a	O
semantically	O
flexible	O
knowledge	O
representation	O
model	O
is	O
needed	O
.	O
</s>
<s>
The	O
main	O
advantage	O
of	O
distributional	O
–	O
relational	O
models	O
over	O
purely	O
logical	O
/	O
semantic	O
web	O
models	O
is	O
the	O
fact	O
that	O
the	O
core	O
semantic	O
associations	O
can	O
be	O
automatically	O
captured	O
from	O
corpora	O
in	O
contrast	O
to	O
the	O
definition	O
of	O
manually	O
curated	O
ontologies	B-Language
and	O
rule	O
knowledge	O
bases	O
.	O
</s>
<s>
Distributional	O
–	O
relational	O
models	O
were	O
first	O
formalized	O
as	O
a	O
mechanism	O
to	O
cope	O
with	O
the	O
vocabulary/semantic	O
gap	O
between	O
users	O
and	O
the	O
schema	B-Application
behind	O
the	O
data	O
.	O
</s>
<s>
In	O
this	O
scenario	O
,	O
distributional	O
semantic	O
relatedness	O
measures	O
,	O
combined	O
with	O
semantic	O
pivoting	O
heuristics	B-Algorithm
can	O
support	O
the	O
approximation	O
between	O
user	O
queries	O
(	O
expressed	O
in	O
their	O
own	O
vocabulary	O
)	O
and	O
data	O
(	O
expressed	O
in	O
the	O
vocabulary	O
of	O
the	O
designer	O
)	O
.	O
</s>
<s>
The	O
geometric	O
aspect	O
supports	O
the	O
semantic	O
approximation	O
between	O
entities	O
from	O
different	O
databases	B-Application
or	O
between	O
a	O
query	O
term	O
and	O
a	O
database	O
entity	O
.	O
</s>
<s>
The	O
distributional	O
relational	O
model	O
then	O
becomes	O
a	O
double	O
layered	O
model	O
where	O
the	O
semantics	O
of	O
the	O
structured	B-Application
data	I-Application
provides	O
the	O
fine-grained	O
semantics	O
intended	O
by	O
the	O
database	B-General_Concept
designer	I-General_Concept
,	O
which	O
is	O
extended	O
by	O
the	O
distributional	O
semantic	O
model	O
which	O
contains	O
the	O
semantic	O
associations	O
expressed	O
at	O
a	O
broader	O
use	O
.	O
</s>
<s>
These	O
models	O
support	O
the	O
generalization	O
from	O
a	O
closed	O
communication	O
scenario	O
(	O
in	O
which	O
database	B-General_Concept
designers	I-General_Concept
and	O
users	O
live	O
in	O
the	O
same	O
context	O
,	O
e.g.	O
</s>
