<s>
ProbLog	B-Application
is	O
a	O
probabilistic	O
logic	O
programming	O
that	O
extends	O
Prolog	B-Language
with	O
probabilities	O
.	O
</s>
<s>
It	O
minimally	O
extends	O
Prolog	B-Language
by	O
adding	O
the	O
notion	O
of	O
a	O
probabilistic	O
fact	O
,	O
which	O
combines	O
the	O
idea	O
of	O
logic	B-Algorithm
atoms	I-Algorithm
and	O
random	O
variables	O
.	O
</s>
<s>
Similarly	O
to	O
Prolog	B-Language
,	O
ProbLog	B-Application
can	O
query	O
an	O
atom	B-Algorithm
.	O
</s>
<s>
While	O
Prolog	B-Language
returns	O
the	O
truth	O
value	O
of	O
the	O
queried	O
atom	B-Algorithm
,	O
ProbLog	B-Application
returns	O
the	O
probability	O
of	O
it	O
being	O
true	O
.	O
</s>
<s>
A	O
probabilistic	O
fact	O
is	O
a	O
pair	O
with	O
an	O
atom	B-Algorithm
and	O
the	O
probability	O
of	O
being	O
true	O
.	O
</s>
<s>
A	O
rule	O
is	O
defined	O
by	O
an	O
atom	B-Algorithm
,	O
called	O
the	O
head	O
,	O
and	O
a	O
finite	O
set	O
of	O
literals	O
,	O
called	O
the	O
body	O
.	O
</s>
<s>
ProbLog	B-Application
programs	O
consist	O
of	O
a	O
set	O
of	O
probabilistic	O
facts	O
and	O
a	O
set	O
of	O
rules	O
.	O
</s>
<s>
Using	O
the	O
distribution	O
semantics	O
,	O
a	O
probability	O
distribution	O
is	O
defined	O
over	O
the	O
two-valued	O
well-founded	B-Application
models	I-Application
of	O
the	O
atoms	O
in	O
the	O
program	O
.	O
</s>
<s>
ProbLog	B-Application
supports	O
multiple	O
tasks	O
:	O
</s>
<s>
ProbLog	B-Application
can	O
for	O
example	O
be	O
used	O
to	O
calculate	O
the	O
probability	O
of	O
getting	O
wet	O
given	O
the	O
probabilities	O
for	O
rain	O
and	O
the	O
probabilities	O
that	O
someone	O
brings	O
an	O
umbrella	O
as	O
follows	O
:	O
</s>
<s>
When	O
ProbLog	B-Application
is	O
asked	O
to	O
solve	O
the	O
"	O
probabilistic	O
inference	O
"	O
task	O
,	O
the	O
query	O
asks	O
for	O
the	O
probability	O
to	O
stay	O
dry	O
on	O
a	O
weekend	O
day	O
.	O
</s>
<s>
When	O
solving	O
the	O
"	O
most	O
probable	O
explanation	O
"	O
task	O
,	O
ProbLog	B-Application
will	O
return	O
the	O
most	O
likely	O
reason	O
for	O
staying	O
dry	O
,	O
i.e.	O
</s>
<s>
The	O
ProbLog	B-Application
language	O
has	O
been	O
implemented	O
as	O
a	O
YAP	O
Prolog	B-Language
library	O
(	O
ProbLog	B-Application
1	O
)	O
.	O
</s>
<s>
The	O
source	O
code	O
of	O
ProbLog	B-Application
2	O
is	O
licensed	O
under	O
Apache	B-Application
License	I-Application
,	I-Application
Version	I-Application
2.0	I-Application
and	O
available	O
on	O
GitHub	B-Application
.	O
</s>
<s>
ProbLog	B-Application
has	O
been	O
extended	O
or	O
used	O
as	O
inspiration	O
for	O
several	O
different	O
variants	O
,	O
including	O
:	O
</s>
<s>
DeepProbLog	O
extends	O
ProbLog	B-Application
by	O
allowing	O
the	O
probability	O
to	O
be	O
parametrized	O
by	O
a	O
neural	O
network	O
.	O
</s>
<s>
DTProblog	O
extends	O
ProbLog	B-Application
with	O
decision	O
theory	O
.	O
</s>
<s>
DC-ProbLog	O
extends	O
ProbLog	B-Application
with	O
distributional	O
facts	O
,	O
meaning	O
that	O
instead	O
of	O
probabilities	O
,	O
a	O
logic	O
atom	B-Algorithm
has	O
a	O
corresponding	O
continuous	O
probability	O
distribution	O
instead	O
.	O
</s>
<s>
aProbLog	O
extends	O
ProbLog	B-Application
by	O
allowing	O
any	O
semiring	O
instead	O
of	O
just	O
probabilities	O
.	O
</s>
<s>
ProbFOIL	O
:	O
given	O
a	O
set	O
of	O
ProbLog	B-Application
facts	O
as	O
a	O
probabilistic	O
relational	O
database	O
,	O
ProbFOIL	O
finds	O
a	O
set	O
of	O
probabilistic	O
rules	O
to	O
predict	O
the	O
facts	O
of	O
one	O
relation	O
based	O
on	O
all	O
other	O
relations	O
.	O
</s>
