NLP for CP
Addressing Constraint Programming with Natural Language Processing
Home
Resources
Publications
Correct
predictions are in
blue
. If we detect only a subset of a labelled sentence, we highlight the caught part as
blue
, the missing part
light blue.
False positives
are in
green
and
false negatives
are in
red
.
Problem Team_Building — Constraint detection
The
HR
department
of
a
company
organizes
an
integration
day
to
welcome
new
employees
.
The
problem
is
to
configure
10
teams
of
6
people
that
respect
the
following
rules
:
There
are
30
new
employees
and
30
existing
employees
.
They
work
in
6
different
services
lettered
A
to
F
.
A
team
must
have
3
existing
employees
and
3
new
employees
,
and
at
most
4
people
from
the
same
service
.
Some
new
employees
are
coached
by
an
existing
employee
,
and
an
existing
employee
can
coach
only
one
new
employee
.
A
new
employee
who
is
coached
must
be
in
the
team
of
his
coach
.
Furthermore
,
employees
of
services
A
and
B
can
not
be
in
the
same
team
;
employees
of
services
E
and
F
can
not
be
in
the
same
team
.
Each
person
is
represented
by
a
number
in
0-59
;
new
employees
are
the
even
numbers
,
existing
employees
are
the
odd
numbers
.
In
service
A
,
the
range
is
0-19
.
In
service
B
,
the
range
is
20-39
.
In
service
C
,
the
range
is
40-44
.
In
service
D
,
the
range
is
45-49
.
In
service
E
,
the
range
is
50-54
.
In
service
F
,
the
range
is
55-59
.
In
Service
A
:
the
couples
coach/coached
new
employee
are
0-1
,
2-3
,
4-5
,
6-7
,
8-9
,
10-11
.
In
Service
B
:
the
couples
coach/coached
new
employee
are
20-21
,
22-23
,
24-25
,
26-27
,
28-29
,
30-31
.
In
Services
C
,
D
,
E
,
F
,
the
couples
coach/coached
new
employee
are
40-41
,
42-43
,
45-46
,
47-48
,
50-51
,
52-53
,
55-56
,
57-58
.
Additional
constraints
:
person
number
5
must
be
with
either
person
41
or
person
51
;
person
number
15
must
be
with
either
40
or
person
51
;
person
number
25
must
be
with
either
40
or
person
50
;
furthermore
,
person
20
is
with
person
24
,
or
person
22
is
with
person
50
.
Problem Team_Building — Detection of the decisions and objects to be modeled
The
HR
department
of
a
company
organizes
an
integration
day
to
welcome
new
employees
.
The
problem
is
to
configure
10
teams
of
6
people
that
respect
the
following
rules
:
There
are
30
new
employees
and
30
existing
employees
.
They
work
in
6
different
services
lettered
A
to
F
.
A
team
must
have
3
existing
employees
and
3
new
employees
,
and
at
most
4
people
from
the
same
service
.
Some
new
employees
are
coached
by
an
existing
employee
,
and
an
existing
employee
can
coach
only
one
new
employee
.
A
new
employee
who
is
coached
must
be
in
the
team
of
his
coach
.
Furthermore
,
employees
of
services
A
and
B
can
not
be
in
the
same
team
;
employees
of
services
E
and
F
can
not
be
in
the
same
team
.
Each
person
is
represented
by
a
number
in
0-59
;
new
employees
are
the
even
numbers
,
existing
employees
are
the
odd
numbers
.
In
service
A
,
the
range
is
0-19
.
In
service
B
,
the
range
is
20-39
.
In
service
C
,
the
range
is
40-44
.
In
service
D
,
the
range
is
45-49
.
In
service
E
,
the
range
is
50-54
.
In
service
F
,
the
range
is
55-59
.
In
Service
A
:
the
couples
coach/coached
new
employee
are
0-1
,
2-3
,
4-5
,
6-7
,
8-9
,
10-11
.
In
Service
B
:
the
couples
coach/coached
new
employee
are
20-21
,
22-23
,
24-25
,
26-27
,
28-29
,
30-31
.
In
Services
C
,
D
,
E
,
F
,
the
couples
coach/coached
new
employee
are
40-41
,
42-43
,
45-46
,
47-48
,
50-51
,
52-53
,
55-56
,
57-58
.
Additional
constraints
:
person
number
5
must
be
with
either
person
41
or
person
51
;
person
number
15
must
be
with
either
40
or
person
51
;
person
number
25
must
be
with
either
40
or
person
50
;
furthermore
,
person
20
is
with
person
24
,
or
person
22
is
with
person
50
.
Back to list