I
nte
rna
t
io
na
l J
o
urna
l o
f
E
lect
rica
l a
nd
Co
m
pu
t
er
E
ng
ineering
(
I
J
E
CE
)
Vo
l.
1
6
,
No
.
3
,
J
u
n
e
20
2
6
,
p
p
.
1407
~
1
4
1
6
I
SS
N:
2088
-
8
7
0
8
,
DOI
: 1
0
.
1
1
5
9
1
/ijece.
v
1
6
i
3
.
pp
1
4
0
7
-
1
4
1
6
1407
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
ec
e.
ia
esco
r
e.
co
m
Retr
iev
a
l
-
a
ug
men
ted
g
enera
tion in
enterpr
ise kno
wle
dg
e
sy
stems:
a
rchit
ect
ure,
b
enefi
ts,
a
nd
a
pplica
tions
M
o
ha
m
m
a
d B
a
qa
r
Ma
s
ter
o
f
C
o
m
p
u
ter
Ap
p
licatio
n
s
Sen
io
r
So
f
twar
e
E
n
g
in
ee
r
,
C
is
co
Sy
s
tem
s
I
n
c,
Un
ited
State
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Sep
1
0
,
2
0
2
5
R
ev
is
ed
Ma
r
1
8
,
2
0
2
6
Acc
ep
ted
Ap
r
2
7
,
2
0
2
6
Th
is
p
a
p
e
r
p
re
se
n
ts
a
n
a
d
a
p
ti
v
e
re
tri
e
v
a
l
-
a
u
g
m
e
n
ted
g
e
n
e
ra
ti
o
n
(RAG
)
fra
m
e
wo
rk
fo
r
e
n
terp
r
ise
k
n
o
wl
e
d
g
e
sy
ste
m
s
th
a
t
c
o
m
b
i
n
e
s
m
u
lt
i
-
so
u
rc
e
in
g
e
stio
n
,
se
m
a
n
ti
c
in
d
e
x
in
g
wit
h
Hu
g
g
in
g
F
a
c
e
e
m
b
e
d
d
in
g
s
a
n
d
F
a
c
e
b
o
o
k
a
rti
ficia
l
in
telli
g
e
n
c
e
sim
il
a
rit
y
se
a
rc
h
(
F
AISS
)
,
m
e
tad
a
ta
-
a
wa
re
re
tri
e
v
a
l,
a
n
d
g
ro
u
n
d
e
d
lar
g
e
lan
g
u
a
g
e
m
o
d
e
l
g
e
n
e
ra
ti
o
n
.
Th
e
re
se
a
rc
h
a
d
d
re
ss
e
s
a
p
e
rsiste
n
t
e
n
ter
p
rise
g
a
p
:
c
ri
ti
c
a
l
k
n
o
wle
d
g
e
is
d
istri
b
u
te
d
a
c
ro
ss
d
o
c
u
m
e
n
tati
o
n
,
ti
c
k
e
ts,
c
o
d
e
re
p
o
sit
o
ries
,
a
n
d
c
o
ll
a
b
o
ra
ti
o
n
t
o
o
ls,
wh
il
e
sta
ti
c
k
e
y
wo
r
d
se
a
rc
h
a
n
d
p
e
ri
o
d
ica
ll
y
re
train
e
d
lan
g
u
a
g
e
m
o
d
e
ls
c
a
n
n
o
t
k
e
e
p
p
a
c
e
with
ra
p
id
l
y
c
h
a
n
g
i
n
g
o
p
e
ra
ti
o
n
a
l
d
a
ta.
Th
e
p
ro
p
o
se
d
a
p
p
ro
a
c
h
c
o
n
tri
b
u
tes
a
p
ri
v
a
c
y
-
p
re
se
rv
i
n
g
a
rc
h
it
e
c
tu
re
,
a
re
tri
e
v
a
l
-
a
n
d
-
fe
e
d
b
a
c
k
l
o
o
p
th
a
t
imp
ro
v
e
s
ra
n
k
i
n
g
q
u
a
li
t
y
o
v
e
r
ti
m
e
,
a
n
d
a
u
n
i
fie
d
wo
rk
f
lo
w
th
a
t
li
n
k
s
e
v
id
e
n
c
e
re
tri
e
v
a
l
to
so
l
u
ti
o
n
r
e
c
o
m
m
e
n
d
a
ti
o
n
.
In
a
n
e
v
a
lu
a
ti
o
n
o
v
e
r
a
1
.
2
m
il
li
o
n
-
d
o
c
u
m
e
n
t
c
o
rp
u
s
a
n
d
a
six
-
we
e
k
p
il
o
t,
th
e
fra
m
e
wo
rk
imp
ro
v
e
d
P
re
c
isio
n
@1
0
fro
m
0
.
5
8
t
o
0
.
8
1
,
re
d
u
c
e
d
d
o
c
u
m
e
n
tatio
n
re
tri
e
v
a
l
late
n
c
y
fro
m
4
5
.
6
s
t
o
1
2
.
3
s,
a
n
d
sh
o
rten
e
d
a
v
e
ra
g
e
b
u
g
-
re
s
o
lu
ti
o
n
ti
m
e
fr
o
m
1
8
.
4
h
to
7
.
2
h
.
Th
e
se
fin
d
in
g
s
in
d
ica
te
th
a
t
e
n
terp
rise
RAG
c
a
n
m
a
teria
ll
y
imp
ro
v
e
tro
u
b
les
h
o
o
ti
n
g
sp
e
e
d
,
k
n
o
wle
d
g
e
re
u
se
,
a
n
d
d
e
c
isio
n
s
u
p
p
o
rt
wh
il
e
m
a
in
tain
in
g
stro
n
g
e
r
c
o
n
tro
l
o
v
e
r
se
n
siti
v
e
o
rg
a
n
iza
ti
o
n
a
l
d
a
ta.
Th
e
b
ro
a
d
e
r
imp
li
c
a
ti
o
n
is
th
a
t
a
d
a
p
ti
v
e
,
g
o
v
e
rn
e
d
RAG
sy
ste
m
s
c
a
n
se
rv
e
a
s
a
p
ra
c
ti
c
a
l
fo
u
n
d
a
ti
o
n
fo
r
f
u
tu
re
e
n
ter
p
ris
e
a
rti
ficia
l
in
telli
g
e
n
c
e
(AI)
a
ss
istan
ts,
a
n
a
ly
ti
c
s p
latf
o
rm
s,
a
n
d
c
o
m
p
li
a
n
c
e
-
a
wa
r
e
d
e
c
isio
n
wo
r
k
flo
ws
.
K
ey
w
o
r
d
s
:
E
n
ter
p
r
is
e
k
n
o
wled
g
e
s
y
s
tem
s
FAI
SS
Feed
b
ac
k
lear
n
in
g
Priv
ac
y
-
p
r
eser
v
i
n
g
AI
R
etr
iev
al
-
au
g
m
en
ted
g
en
er
ati
o
n
Sem
an
tic
r
etr
iev
al
So
lu
tio
n
r
ec
o
m
m
en
d
atio
n
T
h
is i
s
a
n
o
p
e
n
a
c
c
e
ss
a
rticle
u
n
d
e
r th
e
CC B
Y
-
SA
li
c
e
n
se
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
Mo
h
am
m
ad
B
aq
ar
Ma
s
ter
o
f
C
o
m
p
u
ter
Ap
p
licatio
n
s
Sen
io
r
So
f
twar
e
E
n
g
in
ee
r
,
C
is
co
Sy
s
tem
s
I
n
c
San
J
o
s
e,
C
alif
o
r
n
ia,
Un
ited
States
E
m
ail:
b
aq
ar
2
2
@
g
m
ail.
co
m
1.
I
NT
RO
D
UCT
I
O
N
Mo
d
er
n
e
n
ter
p
r
is
es
d
ep
en
d
o
n
p
latf
o
r
m
s
s
u
ch
as
C
o
n
f
lu
e
n
ce
,
J
I
R
A,
Git,
an
d
W
eb
ex
to
ca
p
tu
r
e
d
o
cu
m
e
n
tatio
n
,
in
cid
e
n
ts
,
co
d
e
ev
o
l
u
tio
n
,
an
d
o
p
e
r
atio
n
al
d
is
cu
s
s
io
n
.
Ho
wev
er
,
th
is
k
n
o
wled
g
e
is
f
r
ag
m
en
te
d
ac
r
o
s
s
h
eter
o
g
en
e
o
u
s
s
y
s
tem
s
,
r
ep
r
esen
te
d
in
d
i
f
f
er
en
t
f
o
r
m
ats,
an
d
u
p
d
ated
c
o
n
tin
u
o
u
s
ly
,
wh
ich
m
ak
es
it
d
if
f
icu
lt
f
o
r
e
n
g
in
ee
r
s
an
d
d
ec
is
io
n
-
m
a
k
er
s
to
r
ec
o
v
er
th
e
r
ig
h
t
c
o
n
tex
t
wh
en
tim
e
-
s
en
s
itiv
e
p
r
o
b
lem
s
em
er
g
e.
T
r
ad
itio
n
al
k
ey
wo
r
d
s
ea
r
ch
r
e
m
ain
s
u
s
ef
u
l
f
o
r
lex
ical
lo
o
k
u
p
b
u
t
p
er
f
o
r
m
s
p
o
o
r
ly
w
h
en
th
e
task
r
e
q
u
ir
es
s
em
an
tic
m
atch
in
g
ac
r
o
s
s
is
s
u
e
h
is
to
r
ies,
co
d
e
ch
an
g
es,
an
d
n
a
r
r
ativ
e
d
o
cu
m
e
n
tatio
n
[
1
]
,
[
2
]
.
Dir
ec
t
u
s
e
o
f
lar
g
e
lan
g
u
ag
e
m
o
d
els
(
L
L
Ms)
in
tr
o
d
u
ce
s
a
s
ec
o
n
d
lim
itatio
n
:
h
allu
cin
atio
n
,
wea
k
g
r
o
u
n
d
in
g
in
en
ter
p
r
is
e
-
s
p
ec
if
ic
ev
id
en
ce
,
an
d
u
n
ac
ce
p
tab
le
p
r
iv
ac
y
ex
p
o
s
u
r
e
f
o
r
s
en
s
itiv
e
in
ter
n
al
d
ata
[
3
]
.
Prio
r
wo
r
k
o
n
b
u
g
tr
ia
g
e,
b
u
g
r
ep
o
r
t
an
aly
s
is
,
m
in
in
g
s
o
f
twar
e
r
ep
o
s
ito
r
ies,
an
d
b
u
g
-
f
ix
r
e
co
m
m
en
d
atio
n
also
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
1
6
,
No
.
3
,
J
u
n
e
20
2
6
:
1
4
0
7
-
1
4
1
6
1408
s
h
o
ws
th
e
v
alu
e
o
f
s
o
f
twar
e
-
e
n
g
in
ee
r
in
g
d
ata
m
in
in
g
,
b
u
t
t
h
ese
m
eth
o
d
s
ar
e
ty
p
ically
o
p
tim
ized
f
o
r
n
ar
r
o
w
task
s
r
ath
er
th
an
f
o
r
u
n
if
ied
,
c
o
n
tin
u
o
u
s
ly
u
p
d
ated
en
ter
p
r
is
e
r
ea
s
o
n
in
g
[
4
]
–
[
7
]
.
T
h
is
p
ap
er
ad
d
r
ess
es
th
at
g
ap
th
r
o
u
g
h
an
ad
ap
tiv
e
e
n
ter
p
r
is
e
r
etr
iev
al
-
au
g
m
e
n
ted
g
en
e
r
atio
n
(
R
AG)
f
r
am
ewo
r
k
th
at
co
u
p
les
m
u
lti
-
s
o
u
r
ce
in
g
esti
o
n
,
s
em
an
ti
c
r
etr
iev
al,
m
etad
ata
-
c
o
n
s
tr
a
in
ed
r
an
k
in
g
,
an
d
g
r
o
u
n
d
ed
g
en
e
r
atio
n
with
in
a
p
r
iv
ac
y
-
awa
r
e
d
e
p
lo
y
m
e
n
t
m
o
d
el
[
8
]
–
[
1
0
]
.
T
h
e
n
o
v
elty
o
f
th
e
m
an
u
s
cr
ip
t
lies
in
th
r
ee
in
teg
r
ate
d
c
o
n
tr
ib
u
tio
n
s
.
First,
it
f
o
r
m
alize
s
a
m
o
d
u
lar
e
n
ter
p
r
is
e
ar
ch
ite
ctu
r
e
th
at
u
n
if
ies
d
o
cu
m
e
n
tatio
n
,
tic
k
etin
g
,
co
d
e,
an
d
co
llab
o
r
atio
n
tr
ac
es
in
t
o
a
s
h
ar
ed
r
etr
iev
al
lay
er
r
ath
e
r
th
an
tr
ea
tin
g
th
em
as
is
o
lated
r
ep
o
s
ito
r
ies.
Seco
n
d
,
it
in
t
r
o
d
u
ce
s
an
a
d
ap
tiv
e
f
ee
d
b
ac
k
m
ec
h
a
n
is
m
th
at
u
s
e
s
click
-
th
r
o
u
g
h
a
n
d
r
eso
lu
tio
n
o
u
tco
m
es
to
r
ef
in
e
r
etr
iev
al
q
u
ality
o
v
er
tim
e,
m
ak
in
g
th
e
s
y
s
tem
r
esp
o
n
s
iv
e
to
ch
an
g
in
g
ter
m
in
o
lo
g
y
an
d
u
s
ag
e
p
atter
n
s
.
T
h
ir
d
,
it
ev
alu
ates
th
e
f
r
am
ewo
r
k
n
o
t
o
n
l
y
o
n
r
etr
iev
al
r
e
lev
an
ce
b
u
t
also
o
n
o
p
er
atio
n
al
o
u
tco
m
es
s
u
ch
as
r
eso
lu
tio
n
tim
e,
r
ep
o
r
t
p
r
ep
a
r
atio
n
tim
e,
an
d
u
s
er
ad
o
p
tio
n
.
T
o
g
eth
e
r
,
th
ese
co
n
tr
ib
u
tio
n
s
p
o
s
itio
n
th
e
wo
r
k
b
ey
o
n
d
d
escr
ip
tiv
e
s
y
s
tem
in
teg
r
atio
n
: th
e
p
ap
e
r
d
em
o
n
s
tr
ates h
o
w
en
ter
p
r
is
e
R
AG
ca
n
f
u
n
ctio
n
as
a
g
o
v
er
n
ed
,
e
v
id
en
ce
-
b
ased
d
ec
is
i
o
n
-
s
u
p
p
o
r
t
m
et
h
o
d
th
at
im
p
r
o
v
es
b
o
th
tech
n
ical
tr
o
u
b
lesh
o
o
ti
n
g
an
d
o
r
g
an
izati
o
n
al
lear
n
in
g
.
2.
RE
L
AT
E
D
WO
RK
E
n
ter
p
r
is
e
k
n
o
wled
g
e
ac
ce
s
s
h
as
tr
ad
itio
n
ally
b
ee
n
s
u
p
p
o
r
ted
b
y
k
ey
wo
r
d
-
ce
n
tr
ic
s
ea
r
ch
en
g
in
es
s
u
ch
as
L
u
ce
n
e
an
d
E
last
icSe
ar
ch
,
wh
ich
ar
e
e
f
f
icien
t
f
o
r
ex
ac
t
-
ter
m
lo
o
k
u
p
b
u
t
o
f
ten
u
n
d
er
p
er
f
o
r
m
wh
e
n
en
g
in
ee
r
s
d
escr
ib
e
th
e
s
am
e
is
s
u
e
u
s
in
g
d
if
f
er
e
n
t
v
o
ca
b
u
lar
y
o
r
wh
en
r
elev
an
t
ev
id
en
ce
is
s
p
lit
ac
r
o
s
s
d
o
cu
m
e
n
ts
,
tick
ets,
an
d
co
d
e
ar
tifa
cts
[
1
1
]
,
[
1
2
]
.
Neu
r
al
r
etr
iev
al
m
eth
o
d
s
s
u
ch
as
d
en
s
e
p
ass
ag
e
r
etr
iev
al
(
DPR
)
an
d
tr
an
s
f
o
r
m
e
r
em
b
e
d
d
in
g
s
im
p
r
o
v
e
d
s
em
an
tic
m
a
tch
in
g
[
1
0
]
,
[
1
3
]
,
an
d
R
AG
ex
ten
d
ed
t
h
is
lin
e
o
f
wo
r
k
b
y
co
m
b
in
in
g
r
etr
iev
al
with
g
en
er
atio
n
s
o
th
at
a
n
s
wer
s
ca
n
b
e
g
r
o
u
n
d
e
d
in
r
etr
ie
v
ed
ev
id
en
ce
r
at
h
er
th
an
in
p
a
r
am
etr
ic
m
em
o
r
y
al
o
n
e
[
1
4
]
.
Desp
ite
th
at
p
r
o
g
r
ess
,
m
u
c
h
o
f
th
e
p
u
b
lis
h
ed
R
AG
liter
atu
r
e
tar
g
ets
o
p
en
-
d
o
m
ai
n
q
u
esti
o
n
an
s
wer
in
g
,
d
ialo
g
u
e
,
cu
s
to
m
e
r
s
u
p
p
o
r
t,
o
r
d
o
cu
m
e
n
t
ass
is
t
an
ce
[
1
4
]
–
[
1
7
]
.
E
n
ter
p
r
is
e
d
ep
lo
y
m
en
ts
f
ac
e
a
d
if
f
er
en
t
s
et
o
f
co
n
s
tr
ain
ts
:
p
r
iv
ac
y
b
o
u
n
d
ar
ies,
r
ap
id
l
y
ch
a
n
g
in
g
in
ter
n
al
co
r
p
o
r
a,
d
o
m
ain
-
s
p
ec
if
ic
v
o
ca
b
u
lar
y
,
an
d
th
e
n
ee
d
t
o
co
n
n
ec
t
m
u
ltip
le
o
p
er
atio
n
al
s
y
s
tem
s
in
a
s
in
g
le
r
e
aso
n
in
g
wo
r
k
f
lo
w.
C
o
m
m
er
cial
en
ter
p
r
is
e
ar
tific
i
al
in
tellig
en
ce
(
AI
)
s
tack
s
an
d
v
ec
to
r
-
d
atab
ase
p
ip
elin
es
p
ar
tially
ad
d
r
ess
th
ese
n
ee
d
s
,
b
u
t
th
e
y
o
f
te
n
em
p
h
asi
ze
im
p
lem
en
tatio
n
co
n
v
en
ie
n
c
e
o
v
er
a
c
r
itical
ac
co
u
n
t
o
f
h
o
w
r
etr
iev
al
q
u
ality
,
g
o
v
er
n
an
ce
,
a
n
d
ac
tio
n
a
b
ilit
y
in
ter
ac
t in
r
ea
l o
r
g
an
izatio
n
al
s
ettin
g
s
.
So
f
twar
e
-
en
g
in
ee
r
in
g
r
esear
c
h
p
r
o
v
i
d
es
s
ev
er
al
ad
jace
n
t
b
u
ild
in
g
b
lo
c
k
s
.
B
u
g
L
o
ca
to
r
a
n
d
Dee
p
L
o
c
f
o
cu
s
o
n
b
u
g
lo
ca
lizatio
n
[
1
8
]
,
[
1
9
]
,
b
u
g
tr
ia
g
e
an
d
b
u
g
-
r
e
p
o
r
t
s
tu
d
ies
ex
am
in
e
ass
ig
n
m
en
t
an
d
in
f
o
r
m
atio
n
n
ee
d
s
[
4
]
,
[
5
]
,
m
in
i
n
g
s
o
f
twar
e
r
ep
o
s
ito
r
ies
h
ig
h
lig
h
ts
th
e
s
tr
ateg
ic
v
alu
e
o
f
d
ev
elo
p
m
en
t
t
r
ac
es
[
6
]
,
a
n
d
b
u
g
-
f
ix
r
ec
o
m
m
en
d
atio
n
wo
r
k
d
e
m
o
n
s
tr
ates
h
o
w
h
is
to
r
ical
f
ix
es
ca
n
g
u
id
e
n
ew
r
eso
lu
tio
n
s
[
7
]
.
Ho
wev
e
r
,
th
ese
ap
p
r
o
ac
h
es
ar
e
u
s
u
ally
s
in
g
le
-
p
u
r
p
o
s
e,
tr
ain
e
d
o
n
n
ar
r
o
wer
co
r
p
o
r
a,
o
r
d
esig
n
ed
f
o
r
o
f
f
l
in
e
an
aly
s
is
r
ath
e
r
th
an
f
o
r
a
liv
e
en
ter
p
r
is
e
r
etr
ie
v
al
-
an
d
-
g
en
er
atio
n
lo
o
p
.
T
h
e
f
r
am
ewo
r
k
p
r
o
p
o
s
ed
in
t
h
is
p
ap
er
g
o
es
b
e
y
o
n
d
p
r
io
r
wo
r
k
b
y
c
o
m
b
in
i
n
g
th
ese
s
tr
an
d
s
in
to
a
s
in
g
le
en
ter
p
r
is
e
R
AG
m
eth
o
d
:
it
in
g
ests
m
u
lti
-
s
o
u
r
ce
o
p
e
r
atio
n
al
d
ata,
u
s
es
s
em
an
tic
a
n
d
m
eta
d
ata
-
awa
r
e
r
etr
iev
al
to
ass
em
b
le
ev
id
e
n
c
e,
p
r
eser
v
es
p
r
iv
ac
y
th
r
o
u
g
h
co
n
tr
o
lled
d
ep
l
o
y
m
en
t
b
o
u
n
d
ar
ies,
an
d
clo
s
es
th
e
lo
o
p
with
u
s
er
f
ee
d
b
ac
k
to
im
p
r
o
v
e
r
an
k
in
g
q
u
ality
o
v
e
r
tim
e.
T
h
is
co
m
b
in
atio
n
o
f
ad
ap
tiv
e
r
etr
iev
al,
p
r
iv
ac
y
-
p
r
eser
v
in
g
d
ep
lo
y
m
e
n
t,
an
d
cr
o
s
s
-
s
y
s
tem
r
ea
s
o
n
in
g
co
n
s
titu
tes
th
e
m
an
u
s
cr
ip
t’
s
m
ain
co
n
tr
ib
u
tio
n
an
d
d
if
f
er
e
n
tiates
it
f
r
o
m
b
o
t
h
g
e
n
er
ic
R
AG
im
p
le
m
en
tatio
n
s
an
d
ea
r
lier
task
-
s
p
ec
if
ic
en
ter
p
r
is
e
s
u
p
p
o
r
t
to
o
ls
.
3.
SYST
E
M
DE
SI
G
N
AN
D
R
AG
I
M
P
L
E
M
E
N
T
A
T
I
O
N
T
h
e
p
r
o
p
o
s
ed
s
y
s
tem
is
d
esig
n
ed
as
a
m
o
d
u
lar
,
s
ca
lab
le
ar
ch
itectu
r
e
to
s
ec
u
r
el
y
in
teg
r
ate
h
eter
o
g
en
e
o
u
s
en
ter
p
r
is
e
k
n
o
wled
g
e
s
o
u
r
ce
s
.
I
ts
co
r
e
o
b
jectiv
es
in
clu
d
e
r
etr
iev
in
g
s
em
an
tically
r
elev
an
t
in
f
o
r
m
atio
n
,
g
r
o
u
n
d
in
g
it
with
h
is
to
r
ical
co
n
tex
t,
an
d
d
eliv
er
in
g
ac
tio
n
a
b
le
in
s
ig
h
ts
wh
ile
p
r
o
tectin
g
s
en
s
itiv
e
d
ata
f
r
o
m
ex
ter
n
al
AI
s
er
v
ices
.
T
h
e
h
ig
h
-
le
v
el
ar
ch
itectu
r
e
,
i
llu
s
tr
ated
in
Fig
u
r
e
1
,
d
elin
ea
tes
th
e
in
ter
p
lay
o
f
in
g
esti
o
n
,
in
d
e
x
in
g
,
r
etr
iev
al,
an
d
g
e
n
er
atio
n
c
o
m
p
o
n
en
ts
,
p
r
o
v
id
in
g
a
r
o
b
u
s
t f
r
am
ew
o
r
k
.
3
.
1
.
Da
t
a
ing
estio
n la
y
er
T
h
e
in
g
esti
o
n
p
i
p
elin
e
ag
g
r
e
g
ates
d
iv
er
s
e
en
ter
p
r
is
e
d
ata
th
r
o
u
g
h
API
s
an
d
s
ch
ed
u
le
d
ex
tr
ac
t,
tr
an
s
f
o
r
m
,
lo
a
d
(
E
T
L
)
p
r
o
ce
s
s
es,
en
s
u
r
in
g
co
m
p
r
eh
en
s
iv
e
co
v
er
ag
e
o
f
o
r
g
a
n
izatio
n
al
k
n
o
wled
g
e.
Key
s
o
u
r
ce
s
en
co
m
p
ass
:
a.
C
o
n
f
lu
en
ce
:
I
n
teg
r
ates
d
o
cu
m
en
tatio
n
,
d
esig
n
d
ec
is
io
n
s
,
ar
ch
itectu
r
e
d
iag
r
am
s
,
an
d
tr
o
u
b
lesh
o
o
tin
g
g
u
id
es,
p
r
eser
v
i
n
g
in
s
titu
tio
n
a
l k
n
o
wled
g
e
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
R
etri
ev
a
l
-
a
u
g
men
ted
g
e
n
era
tio
n
in
en
terp
r
is
e
kn
o
w
led
g
e
…
(
Mo
h
a
mma
d
B
a
q
a
r
)
1409
b.
J
I
R
A:
I
n
co
r
p
o
r
ates
p
r
o
ject
tic
k
ets,
in
clu
d
in
g
b
u
g
r
ep
o
r
ts
,
in
cid
en
t
lo
g
s
,
an
d
r
eso
lu
tio
n
n
o
tes,
ca
p
tu
r
in
g
is
s
u
e
h
is
to
r
ies.
c.
Git
r
ep
o
s
ito
r
ies:
E
x
tr
ac
ts
co
m
m
it
m
etad
ata,
d
if
f
s
,
an
d
d
e
v
elo
p
er
co
m
m
e
n
ts
,
co
n
n
ec
tin
g
b
u
g
r
e
p
o
r
ts
to
co
d
e
ev
o
lu
tio
n
.
d.
W
eb
ex
:
Pro
ce
s
s
es
m
ee
tin
g
tr
an
s
cr
ip
ts
an
d
ch
at
lo
g
s
,
r
ev
ea
li
n
g
r
ea
l
-
tim
e
d
is
cu
s
s
io
n
s
o
n
o
u
tag
es
o
r
f
ea
tu
r
e
r
eq
u
ests
.
T
h
e
d
ata
in
g
esti
o
n
la
y
er
ag
g
r
eg
ates
en
ter
p
r
is
e
k
n
o
wled
g
e
f
r
o
m
C
o
n
f
lu
e
n
ce
,
J
I
R
A,
Git
r
ep
o
s
ito
r
ies,
an
d
W
eb
ex
th
r
o
u
g
h
API
s
an
d
s
ch
ed
u
led
E
T
L
jo
b
s
.
R
ath
er
th
an
tr
ea
tin
g
th
ese
s
o
u
r
ce
s
as
s
ep
ar
ate
s
ilo
s
,
th
e
f
r
am
ewo
r
k
n
o
r
m
alize
s
th
em
i
n
to
a
s
h
ar
e
d
s
ch
em
a
co
n
tain
in
g
attr
ib
u
tes
s
u
ch
as
tim
estam
p
,
p
r
o
ject
id
en
tifie
r
,
au
th
o
r
,
ar
tifa
ct
ty
p
e,
an
d
ac
ce
s
s
m
etad
ata.
T
h
is
d
esig
n
ch
o
ice
is
im
p
o
r
tan
t
to
th
e
m
eth
o
d
b
ec
au
s
e
it
en
ab
les
later
r
etr
iev
al
s
tag
es
to
c
o
m
b
i
n
e
s
em
an
tically
s
im
ilar
co
n
te
n
t
with
o
p
er
atio
n
al
f
ilter
s
,
allo
win
g
th
e
s
y
s
tem
to
an
s
wer
n
o
t
o
n
ly
b
r
o
a
d
tr
o
u
b
lesh
o
o
tin
g
q
u
esti
o
n
s
b
u
t
al
s
o
p
r
o
ject
-
o
r
s
ev
er
ity
-
s
p
ec
if
ic
q
u
er
ies
with
o
u
t
r
eb
u
ild
in
g
th
e
in
d
e
x
.
Fig
u
r
e
1
.
Hig
h
-
lev
el
en
ter
p
r
is
e
R
AG
ar
ch
itectu
r
e
s
h
o
win
g
d
ata
in
g
esti
o
n
,
in
d
e
x
in
g
,
r
etr
iev
al,
an
d
g
e
n
er
atio
n
co
m
p
o
n
en
ts
3
.
2
.
Se
m
a
ntic
ind
ex
ing
Fo
llo
win
g
in
g
esti
o
n
,
d
o
c
u
m
e
n
ts
u
n
d
er
g
o
p
r
ep
r
o
ce
s
s
in
g
a
n
d
em
b
ed
d
in
g
g
e
n
er
atio
n
to
f
ac
ilit
ate
ef
f
icien
t r
etr
iev
al.
T
h
is
p
r
o
ce
s
s
in
clu
d
es:
a.
T
ex
t
n
o
r
m
aliza
tio
n
:
I
n
v
o
lv
es
to
k
en
izatio
n
,
s
to
p
wo
r
d
r
em
o
v
al,
a
n
d
an
o
n
y
m
izatio
n
o
f
s
en
s
itiv
e
d
ata
to
en
h
an
ce
p
r
iv
ac
y
a
n
d
d
ata
q
u
al
ity
.
b.
E
m
b
ed
d
in
g
s
:
Gen
er
ates
d
en
s
e
v
ec
to
r
r
e
p
r
esen
tatio
n
s
u
s
in
g
Hu
g
g
in
g
Face
em
b
ed
d
i
n
g
m
o
d
els,
f
in
e
-
tu
n
ed
f
o
r
s
em
an
tic
s
im
ilar
ity
task
s
to
ca
p
tu
r
e
c
o
n
tex
tu
al
n
u
an
ce
s
.
c.
Vec
to
r
i
n
d
e
x
in
g
:
Sto
r
es
em
b
ed
d
in
g
s
in
a
FAI
SS
in
d
ex
o
p
tim
ized
f
o
r
a
p
p
r
o
x
im
ate
n
e
ar
est
n
eig
h
b
o
r
(
ANN)
s
ea
r
ch
,
with
in
d
ices p
a
r
titi
o
n
ed
b
y
p
r
o
ject
co
d
es to
b
o
o
s
t r
etr
iev
al
p
r
ec
is
io
n
.
T
h
is
lay
er
s
u
p
p
o
r
ts
s
ca
lab
ilit
y
,
ac
co
m
m
o
d
atin
g
m
illi
o
n
s
o
f
en
ter
p
r
is
e
r
ec
o
r
d
s
wh
ile
en
s
u
r
in
g
lo
w
-
laten
cy
r
etr
iev
al
(
<1
5
0
m
s
)
,
a
c
r
itical
f
ea
tu
r
e
f
o
r
r
ea
l
-
tim
e
ap
p
licatio
n
s
.
Af
ter
in
g
esti
o
n
,
ea
ch
d
o
cu
m
en
t
is
p
r
ep
r
o
ce
s
s
ed
th
r
o
u
g
h
to
k
en
n
o
r
m
aliza
tio
n
,
s
e
n
s
itiv
e
-
d
ata
an
o
n
y
m
izatio
n
,
a
n
d
ch
u
n
k
in
g
b
ef
o
r
e
s
em
an
tic
em
b
e
d
d
i
n
g
s
ar
e
g
en
er
ate
d
with
h
u
g
g
in
g
f
ac
e
m
o
d
els
o
p
tim
ized
f
o
r
s
im
ilar
ity
s
ea
r
ch
.
T
h
e
r
esu
ltin
g
v
ec
to
r
s
ar
e
s
to
r
ed
in
p
r
o
ject
-
awa
r
e
FAI
SS
in
d
ices
to
s
u
p
p
o
r
t
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
1
6
,
No
.
3
,
J
u
n
e
20
2
6
:
1
4
0
7
-
1
4
1
6
1410
ap
p
r
o
x
im
ate
n
ea
r
est
-
n
eig
h
b
o
r
r
etr
iev
al
at
en
ter
p
r
is
e
s
ca
le.
C
o
n
ce
p
tu
ally
,
th
is
in
d
e
x
in
g
l
ay
er
is
n
o
t
ju
s
t
an
in
f
r
astru
ctu
r
e
d
etail;
it
is
th
e
m
ec
h
an
is
m
th
at
allo
ws
th
e
p
r
o
p
o
s
ed
f
r
am
ewo
r
k
to
p
r
eser
v
e
s
em
an
tic
co
n
tex
t
wh
ile
s
till
m
ee
tin
g
lo
w
-
laten
cy
o
p
er
atio
n
al
r
eq
u
ir
em
en
ts
f
o
r
in
ter
ac
tiv
e
en
ter
p
r
is
e
u
s
e.
3
.
3
.
Ret
rie
v
a
l a
nd
qu
er
y
f
lo
w
Up
o
n
r
ec
eiv
i
n
g
a
u
s
er
q
u
e
r
y
,
t
h
e
s
y
s
tem
ex
ec
u
tes a
s
tr
u
ctu
r
e
d
s
eq
u
en
ce
o
f
o
p
er
atio
n
s
:
a.
Qu
er
y
e
m
b
ed
d
in
g
:
T
r
a
n
s
f
o
r
m
s
th
e
in
p
u
t
in
to
a
d
en
s
e
v
ec
to
r
u
s
in
g
th
e
s
am
e
em
b
ed
d
in
g
m
o
d
el
f
o
r
co
n
s
is
ten
cy
.
b.
FAI
SS
s
ea
r
ch
:
I
d
en
tifie
s
th
e
t
o
p
-
k
s
em
an
tically
clo
s
est
v
ec
t
o
r
s
f
r
o
m
th
e
in
d
ex
,
lev
er
a
g
in
g
ef
f
icien
t
ANN
ca
p
ab
ilit
ies.
c.
Me
tad
ata
f
ilter
in
g
:
Ap
p
lies
f
il
ter
s
(
e.
g
.
,
p
r
o
ject
n
am
e
=
AT
L
,
s
ev
er
ity
=
c
r
itical)
t
o
r
e
f
in
e
th
e
ca
n
d
id
ate
s
et,
en
h
an
cin
g
r
elev
an
ce
.
d.
C
o
n
tex
t
a
s
s
em
b
ly
: Seg
m
en
ts
r
etr
iev
ed
d
o
c
u
m
en
ts
,
r
an
k
s
th
em
b
y
s
em
an
tic
r
elev
an
ce
,
a
n
d
ass
em
b
les th
em
in
to
a
s
tr
u
ctu
r
ed
co
n
tex
t
p
ac
k
a
g
e.
e.
R
AG
l
ay
er
:
Pa
s
s
e
s
th
e
co
n
tex
t
to
an
Op
en
AI
GPT
API
,
g
en
er
atin
g
r
esp
o
n
s
es
g
r
o
u
n
d
ed
in
r
etr
iev
ed
d
ata,
wh
ich
ef
f
ec
tiv
ely
m
itig
ates h
allu
cin
atio
n
s
.
W
h
en
a
u
s
er
s
u
b
m
its
a
q
u
er
y
,
th
e
f
r
am
ew
o
r
k
c
o
n
v
er
ts
it
in
to
th
e
s
am
e
em
b
ed
d
in
g
s
p
ac
e
as
th
e
in
d
ex
ed
a
r
tifa
cts,
r
etr
iev
es
th
e
to
p
-
k
ca
n
d
id
ates
th
r
o
u
g
h
F
AI
SS
s
ea
r
ch
,
an
d
th
en
r
ef
in
e
s
th
at
ca
n
d
id
ate
s
et
with
m
etad
ata
f
ilter
s
s
u
ch
a
s
p
r
o
ject,
s
ev
er
ity
,
o
r
tim
e
win
d
o
w.
T
h
e
r
etr
iev
ed
p
ass
ag
es
ar
e
s
eg
m
en
ted
,
r
an
k
ed
,
a
n
d
ass
em
b
led
in
to
a
co
n
tex
t
p
ac
k
ag
e
th
at
is
p
ass
ed
to
th
e
g
en
er
atio
n
lay
e
r
,
wh
er
e
th
e
lan
g
u
a
g
e
m
o
d
el
is
in
s
tr
u
cted
to
r
em
ain
g
r
o
u
n
d
ed
in
th
e
s
u
p
p
lied
ev
id
en
ce
.
Fig
u
r
e
2
illu
s
tr
ates
th
is
r
etr
iev
al
-
an
d
-
g
en
er
atio
n
wo
r
k
f
lo
w,
i
n
clu
d
in
g
th
e
in
ter
ac
tio
n
b
etwe
en
q
u
er
y
u
n
d
e
r
s
tan
d
in
g
,
d
o
cu
m
en
t
r
etr
iev
al,
an
d
g
r
o
u
n
d
ed
r
ec
o
m
m
en
d
atio
n
g
en
er
atio
n
.
T
h
is
s
eq
u
en
ce
is
ce
n
tr
al
to
t
h
e
p
a
p
er
’
s
m
et
h
o
d
b
ec
au
s
e
it
tu
r
n
s
en
ter
p
r
is
e
R
AG
f
r
o
m
a
g
en
er
i
c
ch
atb
o
t
p
atter
n
in
to
a
co
n
tr
o
lled
r
ea
s
o
n
in
g
p
ip
elin
e
tied
to
au
d
itab
le
ev
id
e
n
ce
.
T
h
is
wo
r
k
f
lo
w
en
s
u
r
es
th
at
all
o
u
tp
u
ts
ar
e
e
v
id
en
ce
-
b
ased
,
alig
n
in
g
with
en
te
r
p
r
i
s
e
d
ata
in
teg
r
ity
r
eq
u
ir
em
e
n
ts
an
d
m
ain
tain
i
n
g
r
eliab
ilit
y
.
Fig
u
r
e
2
.
R
etr
iev
al
an
d
g
en
e
r
a
tio
n
wo
r
k
f
l
o
w
f
o
r
m
etad
ata
-
a
war
e
en
ter
p
r
is
e
R
AG
3
.
4
.
Rew
a
rd
m
ec
ha
nis
m
a
nd
f
ee
db
a
ck
lo
o
p
A
r
ein
f
o
r
ce
m
e
n
t
f
ee
d
b
ac
k
l
o
o
p
en
h
an
ce
s
r
etr
iev
al
q
u
ality
o
v
er
tim
e
th
r
o
u
g
h
t
h
e
f
o
llo
win
g
m
ec
h
an
is
m
s
:
a.
C
lick
-
th
r
o
u
g
h
f
ee
d
b
ac
k
: T
r
ac
k
s
u
s
er
-
s
elec
ted
d
o
cu
m
e
n
ts
to
ass
es
s
r
elev
an
ce
.
b.
R
eso
lu
tio
n
f
ee
d
b
ac
k
: E
v
alu
ate
s
wh
eth
er
s
u
g
g
ested
s
o
lu
tio
n
s
r
ed
u
ce
p
r
o
b
lem
-
s
o
lv
in
g
o
r
r
es
o
lu
tio
n
tim
e.
c.
Ad
ap
tiv
e
r
e
-
r
an
k
i
n
g
:
Up
d
ates
FAI
SS
in
d
ices
with
r
elev
a
n
c
e
s
ig
n
als,
r
ef
in
i
n
g
s
ea
r
c
h
p
e
r
f
o
r
m
an
ce
b
ased
o
n
r
ea
l
-
wo
r
ld
u
s
ag
e
p
atter
n
s
.
T
h
is
iter
ativ
e
p
r
o
ce
s
s
,
g
r
o
u
n
d
ed
in
r
ein
f
o
r
ce
m
en
t
lear
n
i
n
g
p
r
in
cip
les
[
2
0
]
,
en
s
u
r
es
c
o
n
ti
n
u
o
u
s
im
p
r
o
v
em
en
t
in
r
etr
iev
al
ac
cu
r
ac
y
an
d
ad
a
p
tab
ilit
y
.
A
d
is
tin
ctiv
e
f
ea
tu
r
e
o
f
th
e
p
r
o
p
o
s
ed
f
r
a
m
ewo
r
k
is
its
ad
ap
tiv
e
f
ee
d
b
ac
k
lo
o
p
.
User
click
-
th
r
o
u
g
h
b
eh
av
io
r
in
d
icate
s
wh
eth
er
r
etr
iev
ed
d
o
cu
m
en
ts
ap
p
ea
r
r
elev
an
t,
wh
ile
d
o
wn
s
tr
ea
m
r
eso
lu
tio
n
o
u
tco
m
es
s
h
o
w
wh
eth
er
a
r
ec
o
m
m
en
d
e
d
an
s
wer
ac
tu
ally
h
elp
ed
s
h
o
r
ten
tr
o
u
b
lesh
o
o
tin
g
tim
e.
T
h
ese
two
s
ig
n
als
ar
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
R
etri
ev
a
l
-
a
u
g
men
ted
g
e
n
era
tio
n
in
en
terp
r
is
e
kn
o
w
led
g
e
…
(
Mo
h
a
mma
d
B
a
q
a
r
)
1411
u
s
ed
to
ad
ju
s
t
r
a
n
k
in
g
b
eh
av
i
o
r
a
n
d
r
ef
r
esh
th
e
r
etr
iev
al
lay
er
,
c
r
ea
tin
g
a
p
r
ac
tical
r
ein
f
o
r
ce
m
en
t
m
ec
h
an
is
m
f
o
r
e
n
ter
p
r
is
e
e
n
v
ir
o
n
m
en
ts
wh
er
e
ter
m
in
o
lo
g
y
,
p
r
io
r
ities
,
an
d
ar
tifa
ct
d
is
tr
ib
u
tio
n
s
s
h
if
t
o
v
er
tim
e.
T
h
e
f
ee
d
b
ac
k
lo
o
p
th
er
ef
o
r
e
f
u
n
c
tio
n
s
as
a
co
r
e
n
o
v
elty
o
f
th
e
ap
p
r
o
ac
h
r
at
h
er
th
an
as
an
o
p
tio
n
al
u
s
ab
ilit
y
en
h
an
ce
m
e
n
t.
3
.
5
.
User
-
f
o
cus
ed
f
un
ct
io
na
l
it
ies a
nd
s
t
ra
t
eg
ic
ins
ig
hts
T
h
e
s
y
s
tem
d
eliv
er
s
tailo
r
ed
f
u
n
ctio
n
alities
to
en
h
a
n
ce
u
s
er
ex
p
er
ien
ce
a
n
d
s
u
p
p
o
r
t
d
ec
is
io
n
-
m
ak
i
n
g
p
r
o
ce
s
s
es:
a.
B
u
g
s
im
ilar
ity
s
ea
r
ch
:
Ma
tch
e
s
n
ew
er
r
o
r
lo
g
s
(
e.
g
.
,
“SSL
Han
d
s
h
ak
e
Failu
r
e”
)
to
h
is
to
r
i
ca
l
J
I
R
A
tick
ets
with
s
im
ilar
d
escr
ip
tio
n
s
an
d
r
eso
lu
tio
n
s
,
d
r
awin
g
f
r
o
m
v
ar
i
o
u
s
p
r
o
jects.
b.
C
o
d
e
-
ch
an
g
e
s
u
g
g
esti
o
n
s
:
C
o
r
r
elate
s
tick
ets
with
Git
co
m
m
its
to
r
ec
o
m
m
en
d
r
elev
a
n
t
s
o
u
r
ce
f
iles
f
o
r
in
v
esti
g
atio
n
.
c.
Deb
u
g
g
in
g
a
s
s
is
tan
t:
I
n
f
er
s
p
r
o
b
a
b
le
r
o
o
t
ca
u
s
es
b
ased
o
n
p
r
io
r
tr
o
u
b
lesh
o
o
tin
g
p
att
er
n
s
,
im
p
r
o
v
in
g
d
iag
n
o
s
tic
ef
f
icien
cy
.
Fo
r
in
s
tan
ce
,
a
q
u
er
y
r
eg
a
r
d
in
g
a
n
“SSL
h
an
d
s
h
ak
e
f
ailu
r
e
”
c
o
u
ld
p
r
o
v
id
e
ac
ce
s
s
to
1
2
p
r
i
o
r
tick
ets,
in
clu
d
i
n
g
ass
o
ciate
d
co
m
m
its
an
d
co
n
f
i
g
u
r
atio
n
ch
an
g
es,
th
er
eb
y
s
tr
ea
m
lin
in
g
r
eso
lu
tio
n
.
Ad
d
itio
n
ally
,
th
e
s
y
s
tem
o
f
f
er
s
s
tr
ateg
ic
in
s
ig
h
ts
th
r
o
u
g
h
:
a.
Hig
h
-
lev
el
p
r
o
ject
s
u
m
m
ar
ies
:
Pro
v
id
es
AI
-
g
e
n
er
ated
r
ep
o
r
ts
o
n
r
ec
u
r
r
in
g
b
u
g
p
atter
n
s
,
s
u
ch
as
a
4
1
%
p
r
ev
alen
ce
o
f
d
ata
b
ase
s
ch
em
a
m
ig
r
atio
n
is
s
u
es in
AT
L
d
u
r
in
g
Q2
.
b.
R
is
k
i
n
s
ig
h
ts
: I
d
en
tifie
s
tr
en
d
s
in
s
ev
er
ity
,
f
r
eq
u
e
n
cy
,
a
n
d
r
e
s
o
lu
tio
n
tim
e
to
in
f
o
r
m
r
is
k
m
itig
atio
n
.
c.
Pro
d
u
ctiv
ity
d
ash
b
o
ar
d
s
: H
ig
h
lig
h
ts
b
o
ttlen
ec
k
s
an
d
r
ep
etitiv
e
is
s
u
es,
s
u
p
p
o
r
tin
g
r
eso
u
r
ce
o
p
tim
izatio
n
.
T
h
e
u
s
er
-
f
ac
i
n
g
lay
e
r
tr
an
s
l
ates
th
e
tech
n
ical
p
ip
elin
e
in
to
co
n
c
r
ete
en
ter
p
r
is
e
b
en
ef
its
.
Fo
r
en
g
in
ee
r
s
,
th
e
s
y
s
tem
ca
n
s
u
r
f
ac
e
s
im
ilar
b
u
g
s
,
lin
k
tick
ets
to
r
elev
an
t
co
m
m
its
,
an
d
in
f
e
r
lik
ely
r
o
o
t
ca
u
s
es
f
r
o
m
ea
r
lier
tr
o
u
b
lesh
o
o
tin
g
e
p
is
o
d
es.
Fo
r
m
an
ag
er
s
an
d
o
p
er
atio
n
al
lead
s
,
th
e
s
am
e
r
etr
iev
al
b
ase
s
u
p
p
o
r
ts
h
ig
h
er
-
le
v
el
s
u
m
m
ar
ies
o
f
r
ec
u
r
r
in
g
is
s
u
e
ca
teg
o
r
ies,
r
is
k
tr
en
d
s
,
an
d
p
r
o
d
u
ctiv
ity
b
o
ttlen
ec
k
s
.
B
y
s
u
p
p
o
r
tin
g
b
o
th
f
r
o
n
tlin
e
d
eb
u
g
g
in
g
an
d
s
tr
ateg
ic
r
ep
o
r
tin
g
f
r
o
m
th
e
s
am
e
ev
id
en
ce
ch
ain
,
th
e
f
r
am
e
wo
r
k
d
em
o
n
s
tr
ates
b
r
o
ad
e
r
o
r
g
an
izatio
n
al
v
al
u
e
t
h
an
p
r
i
o
r
to
o
ls
th
at
f
o
cu
s
o
n
ly
o
n
s
ea
r
ch
o
r
o
n
ly
o
n
b
u
g
lo
ca
l
izatio
n
.
4.
E
VA
L
UA
T
I
O
N
AND
R
E
SU
L
T
S
T
h
e
ev
alu
atio
n
was
d
esig
n
ed
to
test
wh
eth
er
th
e
p
r
o
p
o
s
ed
f
r
am
ew
o
r
k
im
p
r
o
v
es
en
ter
p
r
is
e
k
n
o
wled
g
e
wo
r
k
in
way
s
th
at
m
atter
o
p
er
atio
n
ally
,
n
o
t
o
n
ly
wh
eth
e
r
it
r
etr
iev
es
s
e
m
an
tically
s
im
ilar
p
ass
ag
es.
W
e
th
er
ef
o
r
e
co
m
p
ar
ed
th
e
R
AG
s
y
s
tem
with
t
h
e
k
ey
wo
r
d
-
b
ased
J
I
R
A
an
d
C
o
n
f
lu
en
ce
s
ea
r
ch
to
o
ls
alr
ea
d
y
u
s
ed
in
p
r
ac
tice
an
d
ass
ess
ed
th
r
ee
lin
k
e
d
o
u
tc
o
m
es:
r
etr
iev
al
r
ele
v
an
ce
,
s
o
lu
tio
n
ac
cu
r
ac
y
,
an
d
tim
e
-
to
-
r
eso
lu
tio
n
.
T
h
is
d
esig
n
alig
n
s
th
e
ev
al
u
atio
n
with
t
h
e
m
an
u
s
cr
ip
t
’
s
co
r
e
claim
th
at
en
ter
p
r
is
e
R
AG
s
h
o
u
ld
b
e
j
u
d
g
e
d
as a
d
ec
is
io
n
-
s
u
p
p
o
r
t
m
eth
o
d
r
ath
er
t
h
an
as
a
s
tan
d
alo
n
e
lan
g
u
ag
e
-
m
o
d
el
d
em
o
n
s
tr
atio
n
.
4
.
1
.
E
x
perim
ent
a
l
s
et
up
T
h
e
ex
p
e
r
im
en
tal
d
ataset
co
n
tain
ed
ap
p
r
o
x
im
ately
1
.
2
m
ill
io
n
ar
tifa
cts
co
llected
o
v
er
s
ev
en
y
ea
r
s
f
r
o
m
C
o
n
f
lu
en
ce
,
J
I
R
A,
Git,
a
n
d
W
eb
ex
,
p
r
o
v
id
in
g
a
r
ea
lis
tic
r
ep
r
esen
tatio
n
o
f
a
lar
g
e
e
n
ter
p
r
is
e
k
n
o
wled
g
e
b
ase.
W
e
ev
alu
ated
th
e
s
y
s
tem
o
n
5
0
0
au
th
e
n
tic
u
s
er
q
u
e
r
i
es
s
p
an
n
in
g
in
cid
e
n
t
d
escr
ip
ti
o
n
s
,
er
r
o
r
m
ess
ag
es,
an
d
tr
o
u
b
lesh
o
o
tin
g
r
eq
u
ests
,
an
d
co
m
p
ar
ed
t
h
e
p
r
o
p
o
s
ed
f
r
am
ewo
r
k
with
th
e
k
ey
wo
r
d
-
s
e
ar
ch
b
aselin
es
u
s
ed
in
d
ay
-
to
-
d
a
y
w
o
r
k
.
T
h
e
an
aly
s
is
f
o
cu
s
ed
o
n
t
op
-
5
r
elev
a
n
ce
ac
cu
r
ac
y
,
h
u
m
an
-
ju
d
g
ed
s
o
lu
tio
n
ac
cu
r
ac
y
,
an
d
av
er
ag
e
tim
e
-
to
-
r
eso
lu
tio
n
,
b
ec
au
s
e
to
g
eth
er
th
ese
m
etr
ic
s
ca
p
tu
r
e
b
o
th
r
etr
ie
v
al
q
u
a
lity
an
d
p
r
ac
tical
u
s
ef
u
ln
ess
.
a.
Data
s
et:
A
co
r
p
u
s
o
f
ap
p
r
o
x
im
ately
1
.
2
m
illi
o
n
d
o
cu
m
e
n
ts
was
as
s
em
b
led
,
in
clu
d
in
g
C
o
n
f
lu
en
ce
p
a
g
es,
J
I
R
A
tick
ets,
Git
co
m
m
its
,
an
d
W
eb
ex
tr
an
s
cr
ip
ts
,
c
o
v
er
i
n
g
th
e
p
ast
s
ev
en
y
ea
r
s
.
T
h
is
d
i
v
er
s
e
co
llectio
n
m
ir
r
o
r
s
th
e
e
n
ter
p
r
is
e’
s
h
is
to
r
i
ca
l k
n
o
wled
g
e
b
ase.
b.
Qu
er
ies:
A
s
am
p
le
o
f
5
0
0
a
u
th
en
tic
u
s
er
q
u
e
r
ies
was
s
elec
ted
,
co
m
p
r
is
in
g
is
s
u
e
d
es
cr
ip
tio
n
s
,
er
r
o
r
m
ess
ag
es,
an
d
tr
o
u
b
lesh
o
o
tin
g
r
eq
u
ests
,
en
s
u
r
in
g
r
ea
l
-
wo
r
ld
r
elev
an
ce
.
c.
B
aselin
es
:
T
h
e
R
A
G
s
y
s
tem
was
b
en
ch
m
ar
k
ed
ag
ain
s
t
th
e
k
ey
wo
r
d
-
b
ased
s
ea
r
ch
f
u
n
ctio
n
alities
o
f
J
I
R
A
an
d
C
o
n
f
lu
e
n
ce
,
r
e
f
lectin
g
s
tan
d
ar
d
o
r
g
a
n
izatio
n
al
p
r
ac
tices
.
d.
E
v
alu
atio
n
m
etr
ics:
−
T
o
p
-
5
r
ele
v
an
ce
ac
cu
r
ac
y
:
T
h
e
p
er
ce
n
tag
e
o
f
q
u
er
ies
wh
er
e
at
least
o
n
e
o
f
th
e
t
op
-
5
r
etr
ie
v
e
d
d
o
cu
m
e
n
ts
co
r
r
esp
o
n
d
ed
t
o
th
e
g
r
o
u
n
d
-
tr
u
th
r
eso
lu
tio
n
,
as e
s
tab
lis
h
ed
in
p
r
io
r
r
esear
ch
[
1
2
]
.
−
So
l
u
ti
o
n
a
c
c
u
r
ac
y
:
E
v
a
lu
ate
d
b
y
ex
p
e
r
i
e
n
ce
d
p
r
o
f
ess
i
o
n
als
t
h
r
o
u
g
h
b
i
n
a
r
y
ju
d
g
m
e
n
ts
(
co
r
r
e
ct/i
n
c
o
r
r
ec
t)
to
ass
ess
w
h
e
th
er
t
h
e
s
y
s
te
m
-
g
en
er
ate
d
s
o
lu
ti
o
n
a
li
g
n
e
d
w
it
h
th
e
v
e
r
i
f
ie
d
r
es
o
l
u
ti
o
n
.
−
T
im
e
-
to
-
r
eso
lu
tio
n
(
T
T
R
)
:
T
h
e
av
er
ag
e
tim
e
(
in
m
in
u
tes
)
r
eq
u
ir
ed
b
y
u
s
er
s
to
id
en
tify
a
r
elev
an
t
s
o
lu
tio
n
d
u
r
i
n
g
s
im
u
lated
p
r
o
b
lem
-
s
o
lv
in
g
s
ess
io
n
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
1
6
,
No
.
3
,
J
u
n
e
20
2
6
:
1
4
0
7
-
1
4
1
6
1412
4
.
2
.
Resul
t
s
a
nd
qu
a
ntit
a
t
iv
e
ev
a
lua
t
io
n
Fig
u
r
e
3
an
d
T
a
b
le
1
s
h
o
w
t
h
at
th
e
R
AG
s
y
s
tem
co
n
s
is
ten
tly
o
u
tp
er
f
o
r
m
ed
th
e
k
e
y
wo
r
d
-
s
ea
r
ch
b
aselin
e
ac
r
o
s
s
all
r
ep
o
r
ted
d
im
en
s
io
n
s
.
T
h
e
s
tr
o
n
g
est
p
atter
n
is
th
at
s
em
an
tic
r
etr
ie
v
al
an
d
g
r
o
u
n
d
e
d
s
y
n
th
esis
im
p
r
o
v
ed
n
o
t
o
n
l
y
d
o
cu
m
e
n
t
r
elev
an
ce
b
u
t
also
d
o
wn
s
tr
ea
m
wo
r
k
q
u
ality
.
Pre
cisi
o
n
g
ain
s
tr
an
s
lated
in
to
m
o
r
e
ac
cu
r
ate
r
ec
o
m
m
en
d
atio
n
s
,
an
d
th
o
s
e
m
o
r
e
ac
cu
r
ate
r
ec
o
m
m
e
n
d
ati
o
n
s
tr
an
s
lated
in
to
f
aster
tr
o
u
b
lesh
o
o
tin
g
a
n
d
less
escalatio
n
.
T
h
is
r
esu
lt
m
atter
s
b
ec
au
s
e
it
s
u
g
g
ests
th
at
ar
c
h
itectu
r
e’
s
b
en
ef
it
is
cu
m
u
lativ
e:
ea
ch
s
tag
e
o
f
th
e
p
ip
elin
e
ad
d
s
v
al
u
e
to
th
e
n
ex
t
r
ath
er
th
a
n
ac
tin
g
as a
n
is
o
lat
ed
o
p
tim
izatio
n
.
Fig
u
r
e
3
.
Qu
a
n
titativ
e
co
m
p
a
r
is
o
n
o
f
b
aselin
e
s
ea
r
ch
a
n
d
th
e
p
r
o
p
o
s
ed
R
AG
s
y
s
tem
ac
r
o
s
s
k
ey
ev
al
u
atio
n
m
etr
ics
Fo
r
r
etr
iev
al
q
u
ality
,
th
e
R
AG
s
y
s
tem
ac
h
iev
ed
8
7
%
t
op
-
5
r
elev
an
ce
ac
cu
r
ac
y
v
e
r
s
u
s
5
8
%
f
o
r
k
ey
wo
r
d
s
ea
r
c
h
.
T
h
is
m
ar
g
in
i
n
d
icate
s
th
at
em
b
ed
d
in
g
-
b
ase
d
r
etr
iev
al
ca
p
tu
r
ed
s
em
an
tica
lly
r
elate
d
in
cid
en
ts
ev
en
wh
e
n
s
u
r
f
ac
e
te
r
m
in
o
lo
g
y
d
if
f
er
ed
,
wh
ich
is
a
co
m
m
o
n
f
ailu
r
e
m
o
d
e
f
o
r
en
ter
p
r
is
e
k
ey
wo
r
d
s
ea
r
ch
.
I
n
p
r
ac
tice,
th
is
m
ea
n
s
en
g
in
ee
r
s
wer
e
m
o
r
e
lik
ely
to
s
ee
h
i
s
to
r
ically
r
elev
an
t
tick
ets,
d
o
cu
m
en
ts
,
an
d
c
o
d
e
ch
an
g
es with
in
th
e
f
ir
s
t f
ew
r
e
s
u
lts
r
ath
er
th
an
h
a
v
in
g
to
r
ef
o
r
m
u
late
q
u
e
r
ies r
ep
ea
ted
ly
.
Fo
r
s
o
lu
tio
n
q
u
ality
,
in
d
e
p
en
d
en
t
ev
alu
ato
r
s
m
ar
k
ed
8
2
%
o
f
R
AG
-
g
en
er
ated
r
ec
o
m
m
e
n
d
atio
n
s
as
co
r
r
ec
t
co
m
p
a
r
ed
with
4
6
%
f
o
r
th
e
b
aselin
e
wo
r
k
f
lo
w.
B
ec
au
s
e
th
e
g
en
er
atio
n
s
tep
was
co
n
s
tr
ain
ed
b
y
r
etr
iev
ed
ev
id
e
n
ce
,
th
e
s
y
s
tem
r
ed
u
ce
d
u
n
s
u
p
p
o
r
ted
s
u
g
g
esti
o
n
s
wh
ile
s
till
s
y
n
th
esizin
g
a
u
s
ab
le
an
s
wer
.
T
h
e
f
in
d
in
g
s
u
p
p
o
r
ts
th
e
p
a
p
er
’
s
c
en
tr
al
claim
th
at
g
r
o
u
n
d
ed
g
en
er
atio
n
is
m
o
r
e
u
s
ef
u
l
th
a
n
ex
p
o
s
in
g
u
s
er
s
to
r
aw
s
ea
r
ch
h
its
alo
n
e.
F
o
r
ef
f
icien
cy
,
u
s
er
s
wo
r
k
in
g
with
th
e
R
AG
s
y
s
tem
r
eso
lv
ed
is
s
u
es
in
an
av
er
ag
e
o
f
1
2
m
in
u
tes
co
m
p
ar
ed
with
3
4
m
in
u
tes
u
n
d
er
k
ey
wo
r
d
s
ea
r
ch
,
a
6
5
%
r
ed
u
ctio
n
.
T
h
e
s
ix
-
wee
k
p
ilo
t
al
s
o
s
h
o
wed
s
h
o
r
ter
r
ep
o
r
t
-
p
r
ep
ar
atio
n
tim
e
an
d
a
h
ig
h
e
r
s
h
a
r
e
o
f
tick
ets
r
eso
lv
ed
with
o
u
t
escalatio
n
.
T
h
es
e
o
p
e
r
atio
n
al
g
ain
s
s
tr
en
g
th
en
th
e
ar
g
u
m
en
t
th
at
en
ter
p
r
is
e
R
AG
s
h
o
u
ld
b
e
u
n
d
er
s
to
o
d
as
a
p
r
o
d
u
ctiv
ity
an
d
c
o
o
r
d
in
atio
n
m
ec
h
an
is
m
,
n
o
t
m
er
ely
as
a
r
etr
iev
al
u
p
g
r
ad
e.
A
s
ix
-
wee
k
d
ep
lo
y
m
e
n
t
with
ap
p
r
o
x
im
ate
ly
1
2
0
u
s
er
s
ac
r
o
s
s
m
u
ltip
le
J
I
R
A
b
o
ar
d
s
f
u
r
th
er
v
alid
ated
th
ese
f
in
d
in
g
s
.
T
a
b
le
1
s
u
m
m
ar
izes
th
e
co
m
p
a
r
ativ
e
o
p
er
atio
n
al
m
etr
ics o
b
s
er
v
ed
d
u
r
in
g
th
e
p
i
lo
t
.
T
ab
le
1
.
C
o
m
p
a
r
ativ
e
o
p
e
r
atio
n
al
p
er
f
o
r
m
an
ce
o
f
k
ey
wo
r
d
s
ea
r
ch
an
d
th
e
p
r
o
p
o
s
ed
R
AG
s
y
s
tem
d
u
r
in
g
th
e
s
ix
-
wee
k
p
ilo
t
M
e
t
r
i
c
K
e
y
w
o
r
d
S
e
a
r
c
h
R
A
G
S
y
s
t
e
m
Δ I
mp
r
o
v
e
me
n
t
A
v
g
.
B
u
g
R
e
s
o
l
u
t
i
o
n
Ti
m
e
(
h
r
s)
1
8
.
4
7
.
2
6
1
%
f
a
s
t
e
r
P
r
e
c
i
s
i
o
n
@
1
0
0
.
5
8
0
.
8
1
+
2
3
%
D
o
c
u
me
n
t
a
t
i
o
n
R
e
t
r
i
e
v
a
l
La
t
e
n
c
y
(
s)
4
5
.
6
1
2
.
3
7
3
%
f
a
s
t
e
r
Le
a
d
e
r
R
e
p
o
r
t
P
r
e
p
Ti
m
e
(
h
r
s)
6
.
5
1
.
2
8
1
%
f
a
s
t
e
r
%
T
i
c
k
e
t
s R
e
s
o
l
v
e
d
w
/
o
Esc
a
l
a
t
i
o
n
5
4
%
8
2
%
+
2
8
%
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
R
etri
ev
a
l
-
a
u
g
men
ted
g
e
n
era
tio
n
in
en
terp
r
is
e
kn
o
w
led
g
e
…
(
Mo
h
a
mma
d
B
a
q
a
r
)
1413
T
h
ese
r
esu
lts
u
n
d
er
s
co
r
e
th
e
s
y
s
tem
’
s
ab
ilit
y
to
im
p
r
o
v
e
k
n
o
wled
g
e
ac
ce
s
s
ib
ilit
y
,
ac
ce
ler
ate
p
r
o
b
lem
r
eso
lu
tio
n
,
a
n
d
en
h
an
ce
d
ec
is
io
n
-
m
ak
in
g
b
y
u
n
co
v
e
r
in
g
cr
o
s
s
-
p
r
o
ject
p
atter
n
s
-
v
a
lid
atin
g
R
AG
as
a
p
r
ac
tical
an
d
im
p
ac
tf
u
l so
lu
tio
n
f
o
r
e
n
ter
p
r
is
e
k
n
o
wled
g
e
s
y
s
tem
s
.
4
.
3
.
User
f
ee
db
a
ck
A
p
o
s
t
-
p
ilo
t
s
u
r
v
ey
o
f
4
5
p
a
r
ticip
an
ts
r
ein
f
o
r
ce
d
t
h
e
q
u
an
tit
ativ
e
f
in
d
in
g
s
.
R
esp
o
n
d
e
n
ts
c
o
n
s
is
ten
tly
r
ep
o
r
ted
th
at
th
e
s
y
s
tem
s
h
o
r
ten
ed
th
e
p
ath
f
r
o
m
an
in
c
o
m
i
n
g
b
u
g
r
e
p
o
r
t
t
o
th
e
s
u
p
p
o
r
tin
g
tick
et,
co
m
m
it,
o
r
d
o
cu
m
e
n
tatio
n
e
v
id
en
ce
n
ee
d
ed
f
o
r
ac
tio
n
.
T
h
ey
also
h
ig
h
li
g
h
ted
t
h
e
u
s
ef
u
ln
ess
o
f
ex
ec
u
tiv
e
s
u
m
m
ar
ies
t
h
at
ex
p
o
s
ed
r
ec
u
r
r
in
g
is
s
u
e
ca
teg
o
r
ies
ac
r
o
s
s
p
r
o
jects,
s
u
g
g
esti
n
g
th
at
th
e
s
am
e
ar
ch
itect
u
r
e
s
u
p
p
o
r
ts
b
o
th
tech
n
ical
tr
iag
e
an
d
m
a
n
ag
er
i
al
o
v
er
s
ig
h
t.
Mo
r
e
t
h
an
7
0
%
o
f
r
esp
o
n
d
en
ts
p
r
ef
e
r
r
ed
th
e
R
AG
wo
r
k
f
lo
w
to
tr
ad
itio
n
al
s
ea
r
ch
,
in
d
icatin
g
th
at
th
e
m
eth
o
d
im
p
r
o
v
e
d
p
er
ce
iv
ed
u
s
ef
u
ln
ess
as
well
as
m
ea
s
u
r
ed
p
er
f
o
r
m
an
ce
.
T
ak
en
t
o
g
eth
er
,
th
e
ev
al
u
atio
n
in
d
icate
s
th
at
s
em
an
tic
r
etr
iev
al
p
lu
s
g
r
o
u
n
d
e
d
g
en
er
atio
n
i
s
ef
f
ec
tiv
e
b
ec
au
s
e
it
alig
n
s
ev
id
en
ce
d
i
s
co
v
er
y
,
r
ec
o
m
m
e
n
d
atio
n
q
u
ality
,
an
d
u
s
er
wo
r
k
f
l
o
w.
At
th
e
s
am
e
tim
e,
th
e
cu
r
r
en
t
s
tu
d
y
r
ep
o
r
ts
ag
g
r
eg
at
e
o
p
er
atio
n
al
o
u
tco
m
es
r
ath
er
th
an
co
n
f
id
en
ce
in
ter
v
als
o
r
p
er
-
q
u
er
y
v
ar
ian
ce
,
s
o
f
u
tu
r
e
r
ep
licatio
n
s
s
h
o
u
ld
e
x
ten
d
t
h
e
ex
p
er
im
en
tal
d
esig
n
with
r
ep
ea
te
d
tr
ials
an
d
d
is
tr
i
b
u
tio
n
al
s
tatis
tics
.
E
v
en
with
th
at
lim
itatio
n
,
th
e
p
r
esen
t
r
esu
lts
p
r
o
v
id
e
s
tr
o
n
g
em
p
ir
ical
ev
id
e
n
ce
th
at
th
e
p
r
o
p
o
s
ed
en
ter
p
r
is
e
R
AG
f
r
am
ewo
r
k
im
p
r
o
v
es r
el
ev
an
ce
,
r
esp
o
n
s
iv
en
ess
,
an
d
u
s
er
tr
u
s
t o
v
er
th
e
b
aselin
e
to
o
l
s
.
5.
CH
AL
L
E
NG
E
S AN
D
L
I
M
I
T
AT
I
O
NS
Alth
o
u
g
h
th
e
R
AG
-
b
ased
s
y
s
tem
d
em
o
n
s
tr
ated
r
o
b
u
s
t
p
er
f
o
r
m
an
ce
,
s
ev
er
al
ch
all
en
g
es
an
d
lim
itatio
n
s
em
er
g
ed
d
u
r
i
n
g
it
s
im
p
lem
en
tatio
n
a
n
d
d
ep
lo
y
m
en
t,
r
e
q
u
ir
in
g
f
u
r
th
er
r
ef
in
e
m
en
t
an
d
s
tr
ateg
ic
co
n
s
id
er
atio
n
.
E
n
ter
p
r
is
e
d
ata
s
o
u
r
ce
s
,
s
u
ch
as
W
eb
ex
tr
an
s
cr
ip
ts
an
d
leg
ac
y
C
o
n
f
l
u
en
ce
p
ag
es,
o
f
ten
co
n
tain
u
n
s
tr
u
ctu
r
ed
,
r
e
d
u
n
d
an
t,
o
r
n
o
is
y
co
n
ten
t,
wh
ich
ca
n
in
ject
ir
r
elev
an
t
co
n
te
x
t
in
to
r
etr
iev
al
p
i
p
elin
es,
d
eg
r
ad
in
g
th
e
g
r
o
u
n
d
in
g
o
f
l
ar
g
e
lan
g
u
ag
e
m
o
d
els
(
L
L
Ms)
an
d
th
e
r
eliab
ilit
y
o
f
g
en
e
r
ated
o
u
tp
u
ts
.
Prio
r
r
esear
ch
h
ig
h
lig
h
ts
th
at
n
o
is
e
in
tr
ain
in
g
an
d
r
etr
iev
al
co
r
p
o
r
a
s
ig
n
if
ican
tly
co
m
p
r
o
m
is
es
th
e
ef
f
ec
tiv
e
n
ess
o
f
LLM
-
b
ased
s
y
s
tem
s
[
2
0
]
,
u
n
d
er
s
co
r
in
g
th
e
n
ee
d
f
o
r
ad
v
a
n
ce
d
p
r
ep
r
o
ce
s
s
in
g
tech
n
iq
u
e
s
.
Ad
d
itio
n
ally
,
th
e
ev
o
lu
tio
n
o
f
p
r
o
ject
-
s
p
ec
if
ic
ter
m
in
o
lo
g
y
an
d
c
o
d
eb
ases
o
v
er
tim
e
lead
s
to
em
b
e
d
d
in
g
d
r
i
f
t,
wh
er
e
h
is
to
r
ically
tr
ain
ed
em
b
e
d
d
in
g
s
f
ail
to
ca
p
tu
r
e
e
m
er
g
in
g
te
ch
n
ical
v
o
ca
b
u
lar
ies,
r
ed
u
cin
g
r
etr
ie
v
al
p
r
ec
is
io
n
.
C
o
n
tin
u
o
u
s
r
e
-
tr
ai
n
in
g
is
ess
en
tial
to
m
itig
ate
th
is
,
y
et
it
in
tr
o
d
u
ce
s
s
u
b
s
tan
tial
co
m
p
u
tatio
n
al
o
v
er
h
ea
d
an
d
o
p
er
atio
n
al
co
m
p
lex
ity
[
2
1
]
,
p
r
esen
tin
g
a
tr
ad
e
-
o
f
f
b
etwe
en
ac
cu
r
ac
y
an
d
r
eso
u
r
ce
ef
f
icien
cy
.
Desp
ite
FAI
S
S
o
p
tim
izatio
n
s
,
th
e
R
AG
f
r
a
m
ewo
r
k
’
s
d
u
al
r
etr
iev
al
a
n
d
g
en
er
atio
n
p
r
o
ce
s
s
es
ad
d
laten
cy
,
n
ec
ess
itatin
g
ca
r
ef
u
l
m
an
a
g
em
en
t
i
n
h
ig
h
-
tr
af
f
ic
en
ter
p
r
is
e
en
v
ir
o
n
m
en
ts
,
wh
ile
s
ca
lin
g
em
b
ed
d
in
g
u
p
d
ates
an
d
L
L
M
q
u
er
ies
ac
r
o
s
s
m
illi
o
n
s
o
f
r
ec
o
r
d
s
r
aises
s
ig
n
if
ican
t
co
s
t
co
n
ce
r
n
s
,
d
em
a
n
d
in
g
co
s
t
-
ef
f
ec
t
iv
e
s
tr
ateg
ies
[
2
2
]
.
E
x
p
lo
r
in
g
au
to
m
ated
n
o
is
e
d
etec
tio
n
an
d
f
ilter
in
g
tech
n
iq
u
es
co
u
ld
f
u
r
th
e
r
en
h
an
ce
d
ata
q
u
ality
an
d
s
tr
ea
m
lin
e
p
r
ep
r
o
ce
s
s
in
g
ef
f
o
r
ts
.
5
.
1
.
K
ey
cha
lleng
es a
nd
m
it
i
g
a
t
io
n str
a
t
e
g
ies
T
ab
le
2
s
u
m
m
ar
izes
th
e
m
ain
d
ep
lo
y
m
en
t
c
h
allen
g
es
o
b
s
er
v
ed
d
u
r
in
g
th
e
p
ilo
t
an
d
th
e
m
itig
atio
n
s
tr
ateg
ies im
p
lied
b
y
th
e
p
r
o
p
o
s
ed
f
r
am
ewo
r
k
.
T
ab
le
2
.
Key
e
n
ter
p
r
is
e
d
e
p
lo
y
m
en
t c
h
allen
g
es a
n
d
m
itig
ati
o
n
s
tr
ateg
ies f
o
r
th
e
p
r
o
p
o
s
ed
R
AG
f
r
am
ewo
r
k
C
h
a
l
l
e
n
g
e
I
mp
a
c
t
M
i
t
i
g
a
t
i
o
n
S
t
r
a
t
e
g
y
C
r
o
ss
-
l
i
n
g
u
a
l
d
o
c
u
me
n
t
a
t
i
o
n
I
n
a
b
i
l
i
t
y
t
o
r
e
t
r
i
e
v
e
/
a
l
i
g
n
k
n
o
w
l
e
d
g
e
a
c
r
o
ss
En
g
l
i
sh
a
n
d
n
o
n
-
En
g
l
i
s
h
d
o
c
u
me
n
t
s
(
e
.
g
.
,
G
e
r
m
a
n
,
H
i
n
d
i
,
Ja
p
a
n
e
se).
I
n
c
o
r
p
o
r
a
t
e
mul
t
i
l
i
n
g
u
a
l
e
mbedd
i
n
g
s
(
e
.
g
.
,
H
u
g
g
i
n
g
F
a
c
e
mu
l
t
i
l
i
n
g
u
a
l
m
o
d
e
l
s)
a
n
d
f
i
n
e
-
t
u
n
e
o
n
e
n
t
e
r
p
r
i
s
e
-
sp
e
c
i
f
i
c
m
u
l
t
i
l
i
n
g
u
a
l
c
o
r
p
o
r
a
.
N
o
i
s
y
m
e
e
t
i
n
g
t
r
a
n
scr
i
p
t
s
(
W
e
b
e
x
)
M
i
s
l
e
a
d
i
n
g
r
e
t
r
i
e
v
a
l
d
u
e
t
o
A
S
R
e
r
r
o
r
s,
f
i
l
l
e
r
w
o
r
d
s,
a
n
d
i
r
r
e
l
e
v
a
n
t
c
o
n
t
e
n
t
.
P
r
e
p
r
o
c
e
ss
i
n
g
w
i
t
h
s
p
e
e
c
h
-
to
-
t
e
x
t
c
l
e
a
n
i
n
g
,
su
mm
a
r
i
z
a
t
i
o
n
,
a
n
d
d
i
sc
o
u
r
s
e
se
g
m
e
n
t
a
t
i
o
n
b
e
f
o
r
e
e
mb
e
d
d
i
n
g
.
Ev
o
l
v
i
n
g
v
o
c
a
b
u
l
a
r
i
e
s
N
e
w
t
e
c
h
n
i
c
a
l
t
e
r
ms
o
r
p
r
o
j
e
c
t
-
s
p
e
c
i
f
i
c
j
a
r
g
o
n
n
o
t
c
a
p
t
u
r
e
d
i
n
e
m
b
e
d
d
i
n
g
s
.
C
o
n
t
i
n
u
o
u
s e
m
b
e
d
d
i
n
g
u
p
d
a
t
e
s
,
i
n
c
r
e
men
t
a
l
f
i
n
e
-
t
u
n
i
n
g
,
a
n
d
g
l
o
ss
a
r
y
-
b
a
s
e
d
r
e
-
r
a
n
k
i
n
g
.
H
a
l
l
u
c
i
n
a
t
i
o
n
r
i
sk
i
n
LL
M
o
u
t
p
u
t
s
I
n
c
o
r
r
e
c
t
o
r
u
n
v
e
r
i
f
i
a
b
l
e
s
o
l
u
t
i
o
n
s s
u
g
g
e
s
t
e
d
t
o
u
s
e
r
s.
S
t
r
i
c
t
g
r
o
u
n
d
i
n
g
w
i
t
h
r
e
t
r
i
e
v
e
d
e
v
i
d
e
n
c
e
,
r
e
t
r
i
e
v
a
l
-
v
e
r
i
f
i
c
a
t
i
o
n
l
o
o
p
,
a
n
d
t
i
c
k
e
t
I
D
/
c
o
mm
i
t
I
D
c
i
t
a
t
i
o
n
e
n
f
o
r
c
e
m
e
n
t
.
S
c
a
l
i
n
g
a
c
r
o
ss
p
r
o
j
e
c
t
s
La
t
e
n
c
y
a
n
d
i
n
d
e
x
si
z
e
i
ss
u
e
s
w
i
t
h
mi
l
l
i
o
n
s
o
f
d
o
c
u
m
e
n
t
s
a
c
r
o
ss m
u
l
t
i
p
l
e
e
n
t
e
r
p
r
i
se
sy
st
e
ms.
P
a
r
t
i
t
i
o
n
e
d
F
A
I
S
S
i
n
d
i
c
e
s,
d
i
st
r
i
b
u
t
e
d
A
N
N
sea
r
c
h
,
a
n
d
c
a
c
h
i
n
g
f
r
e
q
u
e
n
t
l
y
q
u
e
r
i
e
d
c
o
n
t
e
x
t
s.
Ex
p
l
a
i
n
a
b
i
l
i
t
y
a
n
d
t
r
u
st
U
sers h
e
si
t
a
n
t
t
o
r
e
l
y
o
n
A
I
o
u
t
p
u
t
s w
i
t
h
o
u
t
t
r
a
c
e
a
b
i
l
i
t
y
.
A
d
d
e
x
p
l
a
i
n
a
b
i
l
i
t
y
l
a
y
e
r
s (
a
t
t
e
n
t
i
o
n
m
a
p
s,
c
i
t
a
t
i
o
n
t
r
a
c
i
n
g
,
a
n
d
r
e
l
e
v
a
n
c
e
sc
o
r
e
s)
t
o
j
u
s
t
i
f
y
s
y
st
e
m
o
u
t
p
u
t
s.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
1
6
,
No
.
3
,
J
u
n
e
20
2
6
:
1
4
0
7
-
1
4
1
6
1414
Fu
r
th
er
ch
allen
g
es
in
clu
d
e
s
e
cu
r
ity
an
d
co
m
p
lian
ce
is
s
u
es,
as
th
e
s
y
s
tem
’
s
ar
ch
itectu
r
e
p
r
ev
e
n
ts
d
ir
ec
t
ex
p
o
s
u
r
e
o
f
en
ter
p
r
is
e
d
ata
to
ex
ter
n
al
L
L
Ms
b
u
t
in
t
r
o
d
u
ce
s
r
is
k
s
th
r
o
u
g
h
th
ir
d
-
p
a
r
ty
API
in
teg
r
atio
n
,
p
ar
ticu
lar
ly
in
r
eg
u
lated
i
n
d
u
s
tr
ies.
E
m
er
g
in
g
ap
p
r
o
ac
h
es,
s
u
ch
as
o
n
-
p
r
em
is
e
em
b
ed
d
in
g
g
en
e
r
atio
n
an
d
f
in
e
-
tu
n
in
g
o
f
in
-
h
o
u
s
e
L
L
Ms,
ar
e
b
ein
g
ex
p
l
o
r
ed
to
m
ee
t
s
tr
in
g
en
t
r
eg
u
lato
r
y
s
tan
d
ar
d
s
[
2
3
]
.
Mo
r
eo
v
er
,
d
esp
ite
R
AG’
s
s
u
cc
ess
in
r
ed
u
cin
g
h
allu
cin
atio
n
,
u
s
er
s
o
cc
asio
n
ally
q
u
esti
o
n
th
e
in
te
r
p
r
etab
ilit
y
o
f
AI
-
g
en
er
ated
s
u
g
g
esti
o
n
s
,
with
s
tu
d
ies
s
u
g
g
esti
n
g
th
at
p
r
o
v
i
d
in
g
tr
ac
ea
b
le
lin
k
s
b
etwe
en
r
e
tr
iev
ed
d
o
cu
m
e
n
ts
an
d
r
ec
o
m
m
e
n
d
atio
n
s
ca
n
e
n
h
an
ce
tr
u
s
t
in
en
ter
p
r
is
e
AI
s
y
s
tem
s
[
2
4
]
.
I
n
teg
r
atin
g
s
u
ch
ex
p
lain
ab
ilit
y
f
ea
tu
r
es
is
v
ital
f
o
r
b
o
o
s
tin
g
ad
o
p
tio
n
an
d
c
o
n
f
id
e
n
ce
a
m
o
n
g
u
s
er
s
.
T
h
ese
lim
itatio
n
s
co
llectiv
ely
h
ig
h
lig
h
t
ar
ea
s
f
o
r
f
u
tu
r
e
im
p
r
o
v
em
en
t
to
e
n
s
u
r
e
th
e
s
y
s
tem
’
s
s
ca
lab
ilit
y
,
r
eliab
ilit
y
,
a
n
d
p
r
ac
tica
l
u
tili
ty
in
d
iv
er
s
e
en
ter
p
r
is
e
co
n
te
x
ts
.
I
n
v
esti
g
at
in
g
ad
ap
ti
v
e
laten
cy
m
an
ag
e
m
en
t
s
tr
ateg
ies
an
d
c
o
s
t
-
o
p
ti
m
izatio
n
alg
o
r
ith
m
s
co
u
ld
also
ad
d
r
ess
p
er
f
o
r
m
an
c
e
an
d
ec
o
n
o
m
ic
c
o
n
s
tr
ain
ts
in
lar
g
e
-
s
ca
le
d
ep
lo
y
m
en
ts
.
6.
F
UT
UR
E
WO
RK
T
h
e
cu
r
r
en
t
R
AG
-
b
ased
s
y
s
tem
h
as
s
h
o
wca
s
ed
n
o
tab
le
ad
v
an
ce
m
e
n
ts
in
k
n
o
wled
g
e
r
etr
iev
al,
p
r
o
b
lem
-
s
o
lv
in
g
,
a
n
d
s
tr
ateg
ic
in
s
ig
h
ts
,
y
et
s
ev
er
al
p
r
o
m
is
in
g
av
en
u
es
m
er
it
f
u
r
th
e
r
ex
p
lo
r
atio
n
an
d
en
h
an
ce
m
e
n
t.
A
k
ey
d
ir
ec
tio
n
in
v
o
lv
es
ex
p
an
d
i
n
g
th
e
s
y
s
tem
’
s
ca
p
ab
ilit
ies
b
ey
o
n
d
its
cu
r
r
en
t
E
n
g
lis
h
tex
t
-
b
ased
f
o
c
u
s
to
in
clu
d
e
cr
o
s
s
-
lin
g
u
al
co
r
p
o
r
a,
en
ab
li
n
g
m
u
lti
n
atio
n
al
team
s
t
o
q
u
er
y
in
th
e
ir
n
ativ
e
lan
g
u
ag
es
[
2
5
]
.
I
n
teg
r
atin
g
m
u
lti
-
m
o
d
al
s
o
u
r
ce
s
s
u
ch
as
d
iag
r
am
s
,
ar
c
h
itectu
r
e
b
lu
e
p
r
in
ts
,
an
d
lo
g
v
is
u
aliza
tio
n
s
co
u
ld
en
r
ich
r
etr
ie
v
al
o
u
tco
m
es
an
d
im
p
r
o
v
e
th
e
c
o
m
p
r
eh
en
s
iv
en
ess
o
f
AI
-
g
en
er
ated
r
ec
o
m
m
en
d
atio
n
s
[
2
6
]
.
Ad
d
itio
n
ally
,
s
h
if
tin
g
to
f
u
lly
o
n
-
p
r
em
is
e
d
ep
lo
y
m
en
ts
o
f
em
b
ed
d
in
g
m
o
d
els
an
d
lar
g
e
lan
g
u
ag
e
m
o
d
els
(
L
L
Ms)
co
u
ld
ad
d
r
ess
co
m
p
l
ian
ce
r
eq
u
ir
em
en
ts
in
r
eg
u
lated
in
d
u
s
tr
ies
[
2
7
]
,
wh
ile
f
in
e
-
tu
n
in
g
L
L
Ms
o
n
d
o
m
ain
-
s
p
ec
if
ic
c
o
r
p
o
r
a
(
e
.
g
.
,
n
etwo
r
k
i
n
g
o
r
cy
b
er
s
ec
u
r
it
y
)
o
f
f
er
s
p
o
te
n
tial
to
r
e
d
u
ce
h
allu
cin
atio
n
an
d
en
h
an
ce
c
o
n
tex
tu
al
ac
c
u
r
ac
y
[
2
8
]
.
Ad
d
r
ess
in
g
em
b
e
d
d
in
g
d
r
if
t
th
r
o
u
g
h
co
n
tin
u
al
lear
n
in
g
ap
p
r
o
ac
h
es,
wh
ich
d
y
n
am
ically
u
p
d
ate
em
b
ed
d
i
n
g
s
with
o
u
t
f
u
ll
r
et
r
ain
in
g
,
a
n
d
in
co
r
p
o
r
atin
g
r
ein
f
o
r
ce
m
e
n
t
lear
n
in
g
s
ig
n
als
f
r
o
m
u
s
er
in
ter
ac
tio
n
s
co
u
l
d
o
p
tim
ize
r
etr
iev
al
r
an
k
i
n
g
s
o
v
e
r
tim
e
[
2
9
]
,
[
3
0
]
.
Fu
r
th
er
m
o
r
e
,
em
b
e
d
d
in
g
R
AG
in
s
ig
h
ts
in
to
co
n
tin
u
o
u
s
in
teg
r
atio
n
/co
n
tin
u
o
u
s
d
ep
lo
y
m
en
t
(
C
I
/C
D)
p
ip
elin
es
-
s
u
ch
as
i
d
en
tify
in
g
p
o
ten
tial
r
eg
r
ess
io
n
s
d
u
r
in
g
p
u
ll
r
eq
u
es
ts
-
co
u
ld
ev
o
lv
e
th
e
s
y
s
tem
f
r
o
m
a
r
ea
ctiv
e
tr
o
u
b
lesh
o
o
tin
g
to
o
l
to
a
p
r
o
ac
tiv
e
q
u
ality
ass
u
r
an
ce
m
ec
h
a
n
is
m
[
3
1
]
.
E
x
p
lo
r
in
g
r
ea
l
-
tim
e
d
ata
in
teg
r
atio
n
f
r
o
m
em
er
g
in
g
en
ter
p
r
is
e
to
o
ls
c
o
u
ld
f
u
r
th
er
e
n
h
an
ce
its
r
esp
o
n
s
iv
e
n
ess
to
d
y
n
am
ic
wo
r
k
f
lo
ws.
Fu
tu
r
e
en
h
a
n
ce
m
en
ts
s
h
o
u
ld
a
ls
o
em
p
h
asize
h
u
m
an
-
AI
co
ll
ab
o
r
atio
n
an
d
ex
p
lain
a
b
ilit
y
,
i
n
teg
r
atin
g
tr
an
s
p
ar
en
t
lin
k
s
b
etwe
en
r
e
tr
iev
ed
ev
id
e
n
ce
an
d
g
en
er
at
ed
an
s
wer
s
to
b
u
ild
u
s
er
tr
u
s
t
[
3
2
]
,
alo
n
g
s
id
e
in
ter
ac
tiv
e
q
u
er
y
r
ef
in
e
m
en
t
f
ea
tu
r
es
to
b
o
o
s
t
ad
o
p
tio
n
a
m
o
n
g
u
s
er
s
.
L
astl
y
,
th
e
ab
s
en
ce
o
f
s
tan
d
ar
d
ized
b
en
ch
m
ar
k
s
f
o
r
ev
alu
atin
g
R
AG
s
y
s
tem
s
in
en
ter
p
r
is
e
s
et
tin
g
s
r
ep
r
esen
ts
a
s
ig
n
if
ican
t
g
ap
.
E
s
tab
lis
h
in
g
co
m
m
o
n
d
atasets
,
r
o
b
u
s
t
ev
a
lu
atio
n
f
r
am
ewo
r
k
s
,
an
d
r
e
p
r
o
d
u
cib
le
b
aselin
es
will
p
r
o
m
o
te
co
m
p
ar
ab
ilit
y
ac
r
o
s
s
im
p
lem
en
tatio
n
s
an
d
d
r
iv
e
p
r
o
g
r
ess
in
th
is
f
ield
[
3
3
]
.
T
h
ese
in
itiativ
es
co
llectiv
ely
a
im
to
en
h
an
ce
th
e
s
y
s
tem
’
s
s
ca
lab
il
ity
,
ad
ap
tab
ilit
y
,
an
d
p
r
ac
tical
u
tili
ty
ac
r
o
s
s
d
iv
er
s
e
en
ter
p
r
is
e
en
v
ir
o
n
m
en
ts
.
Ad
d
itio
n
ally
,
in
v
esti
g
atin
g
th
e
s
y
s
tem
’
s
p
er
f
o
r
m
a
n
ce
u
n
d
er
v
ar
y
in
g
d
ata
v
o
lu
m
es
an
d
u
s
er
lo
a
d
s
co
u
l
d
p
r
o
v
i
d
e
in
s
ig
h
ts
in
to
its
r
o
b
u
s
tn
ess
.
L
ev
er
ag
in
g
ad
v
a
n
ce
m
en
ts
in
f
ed
e
r
ated
lear
n
in
g
c
o
u
ld
also
en
a
b
le
s
ec
u
r
e,
d
ec
en
tr
alize
d
k
n
o
wled
g
e
s
h
a
r
in
g
ac
r
o
s
s
m
u
l
tip
le
en
ter
p
r
is
e
en
titi
es,
b
r
o
a
d
en
in
g
its
ap
p
licab
ilit
y
.
7.
CO
NCLU
SI
O
N
T
h
is
s
tu
d
y
o
u
tlin
es
th
e
d
esi
g
n
,
im
p
le
m
en
tatio
n
,
an
d
ev
a
lu
atio
n
o
f
a
R
AG
s
y
s
tem
s
p
ec
if
ically
tailo
r
ed
f
o
r
en
ter
p
r
is
e
k
n
o
wle
d
g
e
m
an
ag
em
e
n
t,
s
ea
m
less
ly
in
teg
r
atin
g
h
eter
o
g
en
eo
u
s
d
ata
s
o
u
r
ce
s
s
u
ch
as
C
o
n
f
lu
en
ce
d
o
cu
m
e
n
tatio
n
,
J
I
R
A
is
s
u
e
h
is
to
r
ies,
Git
co
m
m
it
lo
g
s
,
an
d
W
eb
ex
tr
an
s
cr
ip
ts
.
T
h
is
u
n
if
ied
,
s
em
an
tically
en
r
ich
ed
p
latf
o
r
m
o
f
f
e
r
s
a
co
h
esiv
e
s
o
lu
t
io
n
to
f
r
ag
m
en
te
d
o
r
g
an
iza
tio
n
al
k
n
o
wled
g
e,
d
em
o
n
s
tr
atin
g
s
ig
n
if
ican
t
b
e
n
ef
its
f
o
r
u
s
er
s
.
T
h
e
s
y
s
tem
s
u
r
p
ass
ed
tr
ad
itio
n
al
k
ey
wo
r
d
-
b
ased
s
ea
r
ch
m
eth
o
d
s
,
ac
h
iev
in
g
a
6
1
%
r
e
d
u
ctio
n
in
p
r
o
b
lem
r
eso
lu
tio
n
tim
e,
a
2
3
%
im
p
r
o
v
em
en
t
in
r
etr
iev
al
p
r
ec
is
io
n
,
an
d
a
n
8
1
%
d
ec
r
ea
s
e
in
r
e
p
o
r
t
p
r
ep
a
r
atio
n
tim
e.
T
h
ese
q
u
a
n
tifia
b
le
en
h
an
ce
m
e
n
ts
h
ig
h
lig
h
t
R
AG’
s
ef
f
ec
tiv
en
ess
in
ac
ce
ler
atin
g
r
o
o
t
-
ca
u
s
e
an
al
y
s
is
,
d
eliv
er
in
g
ac
tio
n
ab
le
s
o
lu
tio
n
s
,
an
d
p
r
o
v
id
in
g
h
ig
h
-
lev
el
in
s
ig
h
ts
in
to
r
ec
u
r
r
i
n
g
p
r
o
ject
p
atter
n
s
.
B
ey
o
n
d
its
tech
n
ical
s
tr
en
g
th
s
,
th
e
s
y
s
tem
’
s
p
r
ac
tical
v
alu
e
is
ev
id
en
t
in
its
wid
esp
r
ea
d
ad
o
p
tio
n
,
with
u
s
er
s
ap
p
r
ec
iatin
g
its
ab
i
lity
to
co
r
r
elate
is
s
u
es
w
ith
c
o
d
e
ch
an
g
es
an
d
u
tili
ze
AI
-
g
e
n
er
ated
s
u
m
m
ar
ies
f
o
r
s
tr
ateg
ic
p
lan
n
in
g
.
A
k
ey
ad
v
an
tag
e
o
f
its
ar
ch
itectu
r
e
i
s
th
e
p
r
i
o
r
itizatio
n
o
f
d
ata
p
r
i
v
ac
y
a
n
d
s
ec
u
r
ity
,
ac
h
iev
ed
b
y
k
ee
p
i
n
g
s
en
s
itiv
e
en
ter
p
r
is
e
in
f
o
r
m
atio
n
o
n
-
p
r
em
is
e
an
d
av
o
i
d
in
g
d
ir
ec
t
e
x
p
o
s
u
r
e
to
ex
ter
n
al
m
o
d
els.
Fu
tu
r
e
en
h
a
n
ce
m
en
t
s
co
u
ld
en
c
o
m
p
ass
cr
o
s
s
-
lin
g
u
al
s
u
p
p
o
r
t,
i
n
teg
r
atio
n
o
f
m
u
lti
-
m
o
d
al
d
ata,
d
ep
lo
y
m
e
n
t
o
f
d
o
m
ai
n
-
s
p
ec
i
f
ic
L
L
Ms,
an
d
d
ee
p
er
in
te
g
r
atio
n
with
c
o
n
tin
u
o
u
s
in
t
eg
r
atio
n
/co
n
tin
u
o
u
s
d
ep
lo
y
m
e
n
t
(
C
I
/C
D)
p
ip
elin
es.
Pair
ed
with
th
e
estab
li
s
h
m
en
t
o
f
r
o
b
u
s
t
b
en
ch
m
ar
k
s
an
d
im
p
r
o
v
ed
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
R
etri
ev
a
l
-
a
u
g
men
ted
g
e
n
era
tio
n
in
en
terp
r
is
e
kn
o
w
led
g
e
…
(
Mo
h
a
mma
d
B
a
q
a
r
)
1415
ex
p
lain
ab
ilit
y
f
ea
tu
r
es,
th
ese
d
ev
elo
p
m
e
n
ts
ar
e
p
o
is
ed
to
e
n
h
an
ce
tr
u
s
t,
s
ca
lab
ilit
y
,
an
d
ad
o
p
tio
n
o
f
R
AG
s
y
s
tem
s
in
en
ter
p
r
is
e
s
ettin
g
s
.
Ov
er
all,
th
is
r
esear
ch
b
r
i
d
g
e
s
th
e
g
ap
b
etwe
en
d
is
p
ar
ate
k
n
o
wled
g
e
s
y
s
tem
s
an
d
ac
tio
n
ab
le
in
s
ig
h
ts
,
estab
lis
h
in
g
R
AG
as
a
p
r
ac
tical
en
ab
ler
o
f
p
r
o
d
u
ctiv
ity
,
r
eliab
ilit
y
,
an
d
o
r
g
an
izatio
n
al
i
n
tellig
en
ce
in
co
m
p
lex
o
p
er
atio
n
al
en
v
ir
o
n
m
en
ts
.
F
UNDING
I
NF
O
R
M
A
T
I
O
N
Au
th
o
r
s
s
tate
n
o
f
u
n
d
in
g
in
v
o
lv
ed
.
CO
NF
L
I
C
T
O
F
I
N
T
E
R
E
S
T
ST
A
T
E
M
E
NT
Au
th
o
r
s
s
tate
n
o
co
n
f
lict o
f
in
t
er
est.
DATA AV
AI
L
AB
I
L
I
T
Y
Data
av
ailab
ilit
y
is
n
o
t
ap
p
li
ca
b
le
to
th
is
p
ap
er
as
n
o
n
e
w
d
ata
wer
e
cr
ea
ted
o
r
an
aly
ze
d
in
th
is
s
tu
d
y
.
RE
F
E
R
E
NC
E
S
[
1
]
C
.
D
.
M
a
n
n
i
n
g
,
H
.
S
c
h
ü
t
z
e
,
a
n
d
G
.
W
e
i
k
u
r
n
,
F
o
u
n
d
a
t
i
o
n
s
o
f
st
a
t
i
s
t
i
c
a
l
n
a
t
u
r
a
l
l
a
n
g
u
a
g
e
p
ro
c
e
ss
i
n
g
,
v
o
l
.
3
1
,
n
o
.
3
.
C
a
m
b
r
i
d
g
e
,
M
A
:
M
I
T
P
r
e
ss,
2
0
0
2
.
[
2
]
R
.
B
a
e
z
a
-
Y
a
t
e
s
a
n
d
B
.
R
i
b
e
i
r
o
-
N
e
t
o
,
Mo
d
e
r
n
i
n
f
o
rm
a
t
i
o
n
re
t
ri
e
v
a
l
:
t
h
e
c
o
n
c
e
p
t
s
a
n
d
t
e
c
h
n
o
l
o
g
y
b
e
h
i
n
d
se
a
rc
h
,
2
n
d
e
d
.
,
v
o
l
.
4
8
,
n
o
.
1
2
.
B
o
s
t
o
n
,
M
A
:
A
d
d
i
so
n
-
W
e
s
l
e
y
,
2
0
1
1
.
[
3
]
Z.
J
i
e
t
a
l
.
,
“
S
u
r
v
e
y
o
f
h
a
l
l
u
c
i
n
a
t
i
o
n
i
n
n
a
t
u
r
a
l
l
a
n
g
u
a
g
e
g
e
n
e
r
a
t
i
o
n
,
”
AC
M
C
o
m
p
u
t
i
n
g
S
u
rv
e
y
s
,
v
o
l
.
5
5
,
n
o
.
1
2
,
p
p
.
1
–
3
8
,
2
0
2
3
,
d
o
i
:
1
0
.
1
1
4
5
/
3
5
7
1
7
3
0
.
[
4
]
J.
X
u
a
n
,
H
.
Ji
a
n
g
,
Z.
R
e
n
,
J.
Y
a
n
,
a
n
d
Z.
L
u
o
,
“
A
u
t
o
ma
t
i
c
b
u
g
t
r
i
a
g
e
u
s
i
n
g
sem
i
-
su
p
e
r
v
i
s
e
d
t
e
x
t
c
l
a
ss
i
f
i
c
a
t
i
o
n
,
”
i
n
S
E
K
E
2
0
1
0
-
Pro
c
e
e
d
i
n
g
s
o
f
t
h
e
2
2
n
d
I
n
t
e
rn
a
t
i
o
n
a
l
C
o
n
f
e
r
e
n
c
e
o
n
S
o
f
t
w
a
re
En
g
i
n
e
e
ri
n
g
a
n
d
K
n
o
w
l
e
d
g
e
E
n
g
i
n
e
e
ri
n
g
,
2
0
1
0
,
p
p
.
2
0
9
–
2
1
4
.
[
5
]
S
.
B
r
e
u
,
R
.
P
r
e
mr
a
j
,
J.
S
i
l
l
i
t
o
,
a
n
d
T.
Z
i
mm
e
r
m
a
n
n
,
“
I
n
f
o
r
m
a
t
i
o
n
n
e
e
d
s
i
n
b
u
g
r
e
p
o
r
t
s:
I
mp
r
o
v
i
n
g
c
o
o
p
e
r
a
t
i
o
n
b
e
t
w
e
e
n
d
e
v
e
l
o
p
e
r
s
a
n
d
u
sers
,
”
i
n
Pr
o
c
e
e
d
i
n
g
s
o
f
t
h
e
AC
M
C
o
n
f
e
re
n
c
e
o
n
C
o
m
p
u
t
e
r
S
u
p
p
o
rt
e
d
C
o
o
p
e
r
a
t
i
v
e
Wo
r
k
,
C
S
C
W
,
2
0
1
0
,
p
p
.
3
0
1
–
3
1
0
,
d
o
i
:
1
0
.
1
1
4
5
/
1
7
1
8
9
1
8
.
1
7
1
8
9
7
3
.
[
6
]
A
.
E.
H
a
ssa
n
,
“
T
h
e
r
o
a
d
a
h
e
a
d
f
o
r
mi
n
i
n
g
s
o
f
t
w
a
r
e
r
e
p
o
s
i
t
o
r
i
e
s,”
i
n
Pro
c
e
e
d
i
n
g
s
o
f
t
h
e
2
0
0
8
Fro
n
t
i
e
r
s
o
f
S
o
f
t
w
a
re
M
a
i
n
t
e
n
a
n
c
e
,
Fo
S
M
2
0
0
8
,
2
0
0
8
,
p
p
.
4
8
–
5
7
,
d
o
i
:
1
0
.
1
1
0
9
/
F
O
S
M
.
2
0
0
8
.
4
6
5
9
2
4
8
.
[
7
]
J.
Zh
a
n
g
,
X
.
W
a
n
g
,
D
.
H
a
o
,
T.
X
i
e
,
a
n
d
H
.
M
e
i
,
“
R
e
c
o
mm
e
n
d
i
n
g
r
e
l
e
v
a
n
t
b
u
g
f
i
x
e
s
f
o
r
so
f
t
w
a
r
e
p
r
o
j
e
c
t
s,
”
Pr
o
c
e
e
d
i
n
g
s
o
f
I
C
S
E
2
0
2
1
,
2
0
2
1
.
[
8
]
Y
.
G
a
o
e
t
a
l
.
,
“
R
e
t
r
i
e
v
a
l
-
a
u
g
m
e
n
t
e
d
g
e
n
e
r
a
t
i
o
n
f
o
r
l
a
r
g
e
l
a
n
g
u
a
g
e
mo
d
e
l
s
:
A
su
r
v
e
y
,
”
a
rX
i
v
p
re
p
ri
n
t
a
rXi
v
:
2
3
1
2
.
1
0
9
9
7
,
2
0
2
3
,
d
o
i
:
1
0
.
4
8
5
5
0
/
a
r
X
i
v
.
2
3
1
2
.
1
0
9
9
7
.
[
9
]
J.
Jo
h
n
so
n
,
M
.
D
o
u
z
e
,
a
n
d
H
.
Je
g
o
u
,
“
B
i
l
l
i
o
n
-
sc
a
l
e
si
m
i
l
a
r
i
t
y
s
e
a
r
c
h
w
i
t
h
G
P
U
s,”
I
EEE
T
r
a
n
s
a
c
t
i
o
n
s
o
n
Bi
g
D
a
t
a
,
v
o
l
.
7
,
n
o
.
3
,
p
p
.
5
3
5
–
5
4
7
,
2
0
2
1
,
d
o
i
:
1
0
.
1
1
0
9
/
T
B
D
A
TA
.
2
0
1
9
.
2
9
2
1
5
7
2
.
[
1
0
]
T.
W
o
l
f
e
t
a
l
.
,
“
Tr
a
n
sf
o
r
mers
:
s
t
a
t
e
-
of
-
t
h
e
-
a
r
t
n
a
t
u
r
a
l
l
a
n
g
u
a
g
e
p
r
o
c
e
ss
i
n
g
,
”
i
n
Pro
c
e
e
d
i
n
g
s
o
f
t
h
e
2
0
2
0
c
o
n
f
e
re
n
c
e
o
n
e
m
p
i
r
i
c
a
l
m
e
t
h
o
d
s
i
n
n
a
t
u
r
a
l
l
a
n
g
u
a
g
e
p
ro
c
e
ss
i
n
g
:
syst
e
m
d
e
m
o
n
s
t
r
a
t
i
o
n
s
,
2
0
2
0
,
p
p
.
3
8
–
4
5
,
d
o
i
:
1
0
.
1
8
6
5
3
/
v
1
/
2
0
2
0
.
e
m
n
l
p
-
d
e
m
o
s.
6
.
[
1
1
]
M
.
Le
v
e
n
e
,
S
e
a
rc
h
e
n
g
i
n
e
s:
I
n
f
o
rm
a
t
i
o
n
re
t
ri
e
v
a
l
i
n
p
ra
c
t
i
c
e
,
v
o
l
.
5
4
,
n
o
.
5
.
B
o
st
o
n
,
M
A
:
A
d
d
i
s
o
n
-
W
e
sl
e
y
,
2
0
1
1
.
[
1
2
]
E.
H
a
t
c
h
e
r
,
O
.
G
o
sp
o
d
n
e
t
i
c
,
a
n
d
M
.
M
c
C
a
n
d
l
e
ss
,
L
u
c
e
n
e
i
n
a
c
t
i
o
n
,
2
n
d
e
d
.
G
r
e
e
n
w
i
c
h
,
C
T:
M
a
n
n
i
n
g
P
u
b
l
i
c
a
t
i
o
n
s,
2
0
1
0
.
[
1
3
]
V
.
K
a
r
p
u
k
h
i
n
e
t
a
l
.
,
“
D
e
n
s
e
p
a
ss
a
g
e
r
e
t
r
i
e
v
a
l
f
o
r
o
p
e
n
-
d
o
ma
i
n
q
u
e
st
i
o
n
a
n
sw
e
r
i
n
g
,
”
i
n
EM
N
L
P
2
0
2
0
-
2
0
2
0
C
o
n
f
e
re
n
c
e
o
n
Em
p
i
r
i
c
a
l
Me
t
h
o
d
s
i
n
N
a
t
u
ra
l
L
a
n
g
u
a
g
e
Pr
o
c
e
ss
i
n
g
,
Pr
o
c
e
e
d
i
n
g
s
o
f
t
h
e
C
o
n
f
e
re
n
c
e
,
2
0
2
0
,
p
p
.
6
7
6
9
–
6
7
8
1
,
d
o
i
:
1
0
.
1
8
6
5
3
/
v
1
/
2
0
2
0
.
e
mn
l
p
-
mai
n
.
5
5
0
.
[
1
4
]
P
.
Le
w
i
s
e
t
a
l
.
,
“
R
e
t
r
i
e
v
a
l
-
a
u
g
me
n
t
e
d
g
e
n
e
r
a
t
i
o
n
f
o
r
k
n
o
w
l
e
d
g
e
-
i
n
t
e
n
si
v
e
N
LP
t
a
s
k
s,”
i
n
Ad
v
a
n
c
e
s
i
n
N
e
u
ra
l
I
n
f
o
rm
a
t
i
o
n
Pro
c
e
ssi
n
g
S
y
s
t
e
m
s (
N
e
u
rI
PS
)
,
2
0
2
0
,
v
o
l
.
3
3
,
p
p
.
9
4
5
9
–
9
4
7
4
.
[
1
5
]
K
.
S
h
u
st
e
r
,
S
.
P
o
f
f
,
M
.
C
h
e
n
,
D
.
K
i
e
l
a
,
a
n
d
J.
W
e
s
t
o
n
,
“
R
e
t
r
i
e
v
a
l
a
u
g
me
n
t
a
t
i
o
n
r
e
d
u
c
e
s
h
a
l
l
u
c
i
n
a
t
i
o
n
i
n
c
o
n
v
e
r
s
a
t
i
o
n
,
”
Fi
n
d
i
n
g
s
o
f
t
h
e
Ass
o
c
i
a
t
i
o
n
f
o
r
C
o
m
p
u
t
a
t
i
o
n
a
l
L
i
n
g
u
i
st
i
c
s
,
Fi
n
d
i
n
g
s
o
f
A
C
L
:
EM
N
L
P
2
0
2
1
,
p
p
.
3
7
8
4
–
3
8
0
3
,
2
0
2
1
,
d
o
i
:
1
0
.
1
8
6
5
3
/
v
1
/
2
0
2
1
.
f
i
n
d
i
n
g
s
-
e
m
n
l
p
.
3
2
0
.
[
1
6
]
Y
.
F
e
n
g
e
t
a
l
.
,
“
I
mp
r
o
v
i
n
g
c
u
s
t
o
m
e
r
su
p
p
o
r
t
w
i
t
h
r
e
t
r
i
e
v
a
l
-
a
u
g
m
e
n
t
e
d
g
e
n
e
r
a
t
i
o
n
,
”
Pr
o
c
e
e
d
i
n
g
s
o
f
t
h
e
AC
L
2
0
2
2
W
o
rks
h
o
p
s
,
2
0
2
2
.
[
1
7
]
D
.
Y
u
e
t
a
l
.
,
“
R
A
G
-
e
n
h
a
n
c
e
d
d
o
c
u
m
e
n
t
a
n
a
l
y
si
s
f
o
r
e
n
t
e
r
p
r
i
se
a
p
p
l
i
c
a
t
i
o
n
s,”
Pr
o
c
e
e
d
i
n
g
s
o
f
t
h
e
I
E
EE
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
Bi
g
D
a
t
a
(
B
i
g
D
a
t
a
2
0
2
1
)
,
2
0
2
1
.
[
1
8
]
J.
Z
h
o
u
,
H
.
Z
h
a
n
g
,
a
n
d
D
.
Lo
,
“
W
h
e
r
e
sh
o
u
l
d
t
h
e
b
u
g
s
b
e
f
i
x
e
d
?
M
o
r
e
a
c
c
u
r
a
t
e
i
n
f
o
r
mat
i
o
n
r
e
t
r
i
e
v
a
l
-
b
a
s
e
d
b
u
g
l
o
c
a
l
i
z
a
t
i
o
n
b
a
s
e
d
o
n
b
u
g
r
e
p
o
r
t
s,
”
i
n
Pro
c
e
e
d
i
n
g
s
-
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
S
o
f
t
w
a
re
E
n
g
i
n
e
e
r
i
n
g
,
2
0
1
2
,
p
p
.
1
4
–
2
4
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
S
E.
2
0
1
2
.
6
2
2
7
2
1
0
.
[
1
9
]
W
.
La
m,
Z
.
L
i
,
K
.
W
a
n
g
,
a
n
d
L.
Li
u
,
“
D
e
e
p
l
e
a
r
n
i
n
g
f
o
r
b
u
g
l
o
c
a
l
i
z
a
t
i
o
n
,
”
Pro
c
e
e
d
i
n
g
s
o
f
I
C
S
E
2
0
1
7
,
2
0
1
7
.
[
2
0
]
S
.
G
u
r
u
r
a
n
g
a
n
e
t
a
l
.
,
“
D
o
n
’
t
st
o
p
p
r
e
t
r
a
i
n
i
n
g
:
A
d
a
p
t
l
a
n
g
u
a
g
e
mo
d
e
l
s
t
o
d
o
ma
i
n
s
a
n
d
t
a
s
k
s,
”
i
n
Pr
o
c
e
e
d
i
n
g
s
o
f
t
h
e
An
n
u
a
l
Me
e
t
i
n
g
o
f
t
h
e
Asso
c
i
a
t
i
o
n
f
o
r
C
o
m
p
u
t
a
t
i
o
n
a
l
L
i
n
g
u
i
st
i
c
s
,
2
0
2
0
,
p
p
.
8
3
4
2
–
8
3
6
0
,
d
o
i
:
1
0
.
1
8
6
5
3
/
v
1
/
2
0
2
0
.
a
c
l
-
ma
i
n
.
7
4
0
.
[
2
1
]
A
.
La
z
a
r
i
d
o
u
e
t
a
l
.
,
“
P
i
t
f
a
l
l
s
o
f
st
a
t
i
c
e
mb
e
d
d
i
n
g
s
i
n
d
y
n
a
m
i
c
e
n
v
i
r
o
n
me
n
t
s
,
”
a
rX
i
v
p
re
p
r
i
n
t
a
rX
i
v
:
2
1
0
4
.
0
8
5
4
7
,
2
0
2
1
.
[
2
2
]
J.
K
a
p
l
a
n
e
t
a
l
.
,
“
S
c
a
l
i
n
g
l
a
w
s fo
r
n
e
u
r
a
l
l
a
n
g
u
a
g
e
m
o
d
e
l
s,
”
O
p
e
n
A
I
Te
c
h
n
i
c
a
l
R
e
p
o
r
t
,
2
0
2
0
.
[
2
3
]
E.
B
o
mm
a
s
a
n
i
,
R
.
,
H
u
d
s
o
n
,
D
.
,
A
d
e
l
i
,
“
O
n
t
h
e
o
p
p
o
r
t
u
n
i
t
i
e
s
a
n
d
r
i
s
k
s
o
f
f
o
u
n
d
a
t
i
o
n
m
o
d
e
l
s,”
a
rX
i
v
p
r
e
p
ri
n
t
a
r
Xi
v
:
2
1
0
8
.
0
7
2
5
8
,
2
0
2
1
.
[
2
4
]
M
.
T.
R
i
b
e
i
r
o
,
S
.
S
i
n
g
h
,
a
n
d
C
.
G
u
e
s
t
r
i
n
,
“
‘
W
h
y
sh
o
u
l
d
i
t
r
u
st
y
o
u
?
’
E
x
p
l
a
i
n
i
n
g
t
h
e
p
r
e
d
i
c
t
i
o
n
s
o
f
a
n
y
c
l
a
ss
i
f
i
e
r
,
”
i
n
N
AAC
L
-
H
L
T
2
0
1
6
-
2
0
1
6
C
o
n
f
e
re
n
c
e
o
f
t
h
e
N
o
rt
h
Am
e
ri
c
a
n
C
h
a
p
t
e
r
o
f
t
h
e
Ass
o
c
i
a
t
i
o
n
f
o
r
C
o
m
p
u
t
a
t
i
o
n
a
l
L
i
n
g
u
i
s
t
i
c
s:
H
u
m
a
n
L
a
n
g
u
a
g
e
T
e
c
h
n
o
l
o
g
i
e
s
,
Pr
o
c
e
e
d
i
n
g
s
o
f
t
h
e
D
e
m
o
n
s
t
ra
t
i
o
n
s
S
e
ss
i
o
n
,
2
0
1
6
,
p
p
.
9
7
–
1
0
1
,
d
o
i
:
1
0
.
1
8
6
5
3
/
v
1
/
n
1
6
-
3
0
2
0
.
[
2
5
]
A
.
C
o
n
n
e
a
u
e
t
a
l
.
,
“
U
n
su
p
e
r
v
i
s
e
d
c
r
o
ss
-
l
i
n
g
u
a
l
r
e
p
r
e
se
n
t
a
t
i
o
n
l
e
a
r
n
i
n
g
a
t
sca
l
e
,
”
i
n
Pr
o
c
e
e
d
i
n
g
s
o
f
t
h
e
A
n
n
u
a
l
M
e
e
t
i
n
g
o
f
t
h
e
Asso
c
i
a
t
i
o
n
f
o
r
C
o
m
p
u
t
a
t
i
o
n
a
l
L
i
n
g
u
i
st
i
c
s
,
2
0
2
0
,
p
p
.
8
4
4
0
–
8
4
5
1
,
d
o
i
:
1
0
.
1
8
6
5
3
/
v
1
/
2
0
2
0
.
a
c
l
-
mai
n
.
7
4
7
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
1
6
,
No
.
3
,
J
u
n
e
20
2
6
:
1
4
0
7
-
1
4
1
6
1416
[
2
6
]
X
.
Li
e
t
a
l
.
,
“
M
u
l
t
i
-
m
o
d
a
l
r
e
t
r
i
e
v
a
l
-
a
u
g
me
n
t
e
d
g
e
n
e
r
a
t
i
o
n
f
o
r
k
n
o
w
l
e
d
g
e
-
i
n
t
e
n
si
v
e
t
a
sk
s
,
”
a
rX
i
v
p
re
p
ri
n
t
a
r
Xi
v
:
2
3
0
4
.
0
3
2
7
7
,
2
0
2
3
.
[
2
7
]
H
.
To
u
v
r
o
n
e
t
a
l
.
,
“
L
La
M
A
:
O
p
e
n
a
n
d
e
f
f
i
c
i
e
n
t
f
o
u
n
d
a
t
i
o
n
l
a
n
g
u
a
g
e
mo
d
e
l
s,”
a
rX
i
v
p
re
p
r
i
n
t
a
rX
i
v
:
2
3
0
2
.
1
3
9
7
1
.
2
0
2
3
.
[
2
8
]
S
.
G
u
r
u
r
a
n
g
a
n
e
t
a
l
.
,
“
To
w
a
r
d
s
d
o
ma
i
n
-
sp
e
c
i
f
i
c
l
a
n
g
u
a
g
e
m
o
d
e
l
s,”
Pr
o
c
e
e
d
i
n
g
s
o
f
t
h
e
AC
L
2
0
2
2
W
o
rks
h
o
p
s
,
2
0
2
2
.
[
2
9
]
Z.
C
h
e
n
a
n
d
B
.
L
i
u
,
L
i
f
e
l
o
n
g
m
a
c
h
i
n
e
l
e
a
r
n
i
n
g
,
2
n
d
e
d
.
S
a
n
R
a
f
a
e
l
,
C
A
:
M
o
r
g
a
n
&
C
l
a
y
p
o
o
l
,
2
0
1
8
.
[
3
0
]
S
.
N
i
u
,
X
.
P
a
n
,
J.
W
a
n
g
,
a
n
d
G
.
Li
,
“
D
e
e
p
r
e
i
n
f
o
r
c
e
me
n
t
l
e
a
r
n
i
n
g
f
r
o
m
h
u
ma
n
p
r
e
f
e
r
e
n
c
e
s
f
o
r
R
O
V
p
a
t
h
t
r
a
c
k
i
n
g
,
”
i
n
O
c
e
a
n
En
g
i
n
e
e
ri
n
g
,
2
0
2
5
,
v
o
l
.
3
1
7
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
o
c
e
a
n
e
n
g
.
2
0
2
4
.
1
2
0
0
3
6
.
[
3
1
]
M
.
H
i
l
t
o
n
,
T
.
Tu
n
n
e
l
l
,
K
.
H
u
a
n
g
,
D
.
M
a
r
i
n
o
v
,
a
n
d
D
.
D
i
g
,
“
U
sa
g
e
,
c
o
s
t
s,
a
n
d
b
e
n
e
f
i
t
s
o
f
c
o
n
t
i
n
u
o
u
s
i
n
t
e
g
r
a
t
i
o
n
i
n
o
p
e
n
-
s
o
u
r
c
e
p
r
o
j
e
c
t
s,
”
i
n
AS
E
2
0
1
6
-
Pro
c
e
e
d
i
n
g
s
o
f
t
h
e
3
1
st
I
EEE
/
A
C
M
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
A
u
t
o
m
a
t
e
d
S
o
f
t
w
a
re
E
n
g
i
n
e
e
r
i
n
g
,
2
0
1
6
,
p
p
.
4
2
6
–
4
3
7
,
d
o
i
:
1
0
.
1
1
4
5
/
2
9
7
0
2
7
6
.
2
9
7
0
3
5
8
.
[
3
2
]
Z.
C
.
Li
p
t
o
n
,
“
T
h
e
my
t
h
o
s
o
f
m
o
d
e
l
i
n
t
e
r
p
r
e
t
a
b
i
l
i
t
y
,
”
i
n
C
o
m
m
u
n
i
c
a
t
i
o
n
s
o
f
t
h
e
A
C
M
,
2
0
1
8
,
v
o
l
.
6
1
,
n
o
.
1
0
,
p
p
.
3
5
–
4
3
,
d
o
i
:
1
0
.
1
1
4
5
/
3
2
3
3
2
3
1
.
[
3
3
]
V
.
K
a
r
p
u
k
h
i
n
e
t
a
l
.
,
“
B
e
n
c
h
mar
k
i
n
g
d
e
n
se
r
e
t
r
i
e
v
a
l
f
o
r
e
n
t
e
r
p
r
i
s
e
a
n
d
k
n
o
w
l
e
d
g
e
-
i
n
t
e
n
si
v
e
t
a
sk
s,”
a
r
Xi
v
p
re
p
r
i
n
t
a
rXi
v
:
2
2
1
0
.
0
9
1
6
7
,
2
0
2
2
.
B
I
O
G
RAP
H
I
E
S O
F
AUTH
O
RS
Mo
h
a
m
m
a
d
Ba
q
a
r
M
a
ste
r
o
f
Co
m
p
u
ter
Ap
p
li
c
a
ti
o
n
s
,
S
e
n
io
r
S
o
ftwa
re
En
g
in
e
e
r
a
t
Cisc
o
S
y
ste
m
s,
In
c
,
USA.
He
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
b
a
q
a
r
2
2
@g
m
a
il
.
c
o
m
.
Evaluation Warning : The document was created with Spire.PDF for Python.