mirror of
https://github.com/explosion/spaCy.git
synced 2024-09-22 11:59:14 +03:00
eae35e9b27
- add files to specify rules for German tokenization - change generate_specials.py to generate from an external file (abbrev.de.tab) - copy gazetteer.json from lang_data/en/ - init_model.py - change doc freq threshold to 0 - add train_german_tagger.py - expects conll09-formatted input
320 lines
3.3 KiB
SQL
320 lines
3.3 KiB
SQL
# surface form lemma pos
|
||
# multiple values are separated by |
|
||
# empty lines and lines starting with # are being ignored
|
||
|
||
'' ''
|
||
\") \")
|
||
\n \n <nl> SP
|
||
\t \t <tab> SP
|
||
<space> SP
|
||
|
||
# example: Wie geht's?
|
||
's 's es
|
||
'S 'S es
|
||
|
||
# example: Haste mal 'nen Euro?
|
||
'n 'n ein
|
||
'ne 'ne eine
|
||
'nen 'nen einen
|
||
|
||
# example: Kommen S’ nur herein!
|
||
s' s' sie
|
||
S' S' sie
|
||
|
||
# example: Da haben wir's!
|
||
ich's ich|'s ich|es
|
||
du's du|'s du|es
|
||
er's er|'s er|es
|
||
sie's sie|'s sie|es
|
||
wir's wir|'s wir|es
|
||
ihr's ihr|'s ihr|es
|
||
|
||
# example: Die katze auf'm dach.
|
||
auf'm auf|'m auf|dem
|
||
unter'm unter|'m unter|dem
|
||
über'm über|'m über|dem
|
||
vor'm vor|'m vor|dem
|
||
hinter'm hinter|'m hinter|dem
|
||
|
||
# persons
|
||
B.A. B.A.
|
||
B.Sc. B.Sc.
|
||
Dipl. Dipl.
|
||
Dipl.-Ing. Dipl.-Ing.
|
||
Dr. Dr.
|
||
Fr. Fr.
|
||
Frl. Frl.
|
||
Hr. Hr.
|
||
Hrn. Hrn.
|
||
Frl. Frl.
|
||
Prof. Prof.
|
||
St. St.
|
||
Hrgs. Hrgs.
|
||
Hg. Hg.
|
||
a.Z. a.Z.
|
||
a.D. a.D.
|
||
h.c. h.c.
|
||
Jr. Jr.
|
||
jr. jr.
|
||
jun. jun.
|
||
sen. sen.
|
||
rer. rer.
|
||
Ing. Ing.
|
||
M.A. M.A.
|
||
Mr. Mr.
|
||
M.Sc. M.Sc.
|
||
nat. nat.
|
||
phil. phil.
|
||
|
||
# companies
|
||
Co. Co.
|
||
co. co.
|
||
Cie. Cie.
|
||
A.G. A.G.
|
||
G.m.b.H. G.m.b.H.
|
||
i.G. i.G.
|
||
e.V. e.V.
|
||
|
||
# popular german abbreviations
|
||
Abb. Abb.
|
||
Abk. Abk.
|
||
Abs. Abs.
|
||
Abt. Abt.
|
||
abzgl. abzgl.
|
||
allg. allg.
|
||
a.M. a.M.
|
||
Bd. Bd.
|
||
betr. betr.
|
||
Betr. Betr.
|
||
Biol. Biol.
|
||
biol. biol.
|
||
Bf. Bf.
|
||
Bhf. Bhf.
|
||
Bsp. Bsp.
|
||
bspw. bspw.
|
||
bzgl. bzgl.
|
||
bzw. bzw.
|
||
d.h. d.h.
|
||
dgl. dgl.
|
||
ebd. ebd.
|
||
ehem. ehem.
|
||
eigtl. eigtl.
|
||
entspr. entspr.
|
||
erm. erm.
|
||
ev. ev.
|
||
evtl. evtl.
|
||
Fa. Fa.
|
||
Fam. Fam.
|
||
geb. geb.
|
||
Gebr. Gebr.
|
||
gem. gem.
|
||
ggf. ggf.
|
||
ggü. ggü.
|
||
ggfs. ggfs.
|
||
gegr. gegr.
|
||
Hbf. Hbf.
|
||
Hrsg. Hrsg.
|
||
hrsg. hrsg.
|
||
i.A. i.A.
|
||
i.d.R. i.d.R.
|
||
inkl. inkl.
|
||
insb. insb.
|
||
i.O. i.O.
|
||
i.Tr. i.Tr.
|
||
i.V. i.V.
|
||
jur. jur.
|
||
kath. kath.
|
||
K.O. K.O.
|
||
lt. lt.
|
||
max. max.
|
||
m.E. m.E.
|
||
m.M. m.M.
|
||
mtl. mtl.
|
||
min. min.
|
||
mind. mind.
|
||
MwSt. MwSt.
|
||
Nr. Nr.
|
||
o.a. o.a.
|
||
o.ä. o.ä.
|
||
o.Ä. o.Ä.
|
||
o.g. o.g.
|
||
o.k. o.k.
|
||
O.K. O.K.
|
||
Orig. Orig.
|
||
orig. orig.
|
||
pers. pers.
|
||
Pkt. Pkt.
|
||
Red. Red.
|
||
röm. röm.
|
||
s.o. s.o.
|
||
sog. sog.
|
||
std. std.
|
||
stellv. stellv.
|
||
Str. Str.
|
||
tägl. tägl.
|
||
Tel. Tel.
|
||
u.a. u.a.
|
||
usf. usf.
|
||
u.s.w. u.s.w.
|
||
usw. usw.
|
||
u.U. u.U.
|
||
u.v.m. u.v.m.
|
||
uvm. uvm.
|
||
v.a. v.a.
|
||
vgl. vgl.
|
||
vllt. vllt.
|
||
v.l.n.r. v.l.n.r.
|
||
vlt. vlt.
|
||
Vol. Vol.
|
||
wiss. wiss.
|
||
Univ. Univ.
|
||
z.B. z.B.
|
||
z.b. z.b.
|
||
z.Bsp. z.Bsp.
|
||
z.T. z.T.
|
||
z.Z. z.Z.
|
||
zzgl. zzgl.
|
||
z.Zt. z.Zt.
|
||
|
||
# popular latin abbreviations
|
||
vs. vs.
|
||
adv. adv.
|
||
Chr. Chr.
|
||
A.C. A.C.
|
||
A.D. A.D.
|
||
e.g. e.g.
|
||
i.e. i.e.
|
||
al. al.
|
||
p.a. p.a.
|
||
P.S. P.S.
|
||
q.e.d. q.e.d.
|
||
R.I.P. R.I.P.
|
||
etc. etc.
|
||
incl. incl.
|
||
ca. ca.
|
||
n.Chr. n.Chr.
|
||
p.s. p.s.
|
||
v.Chr. v.Chr.
|
||
|
||
# popular english abbreviations
|
||
D.C. D.C.
|
||
N.Y. N.Y.
|
||
N.Y.C. N.Y.C.
|
||
U.S. U.S.
|
||
U.S.A. U.S.A.
|
||
L.A. L.A.
|
||
U.S.S. U.S.S.
|
||
|
||
# dates & time
|
||
Jan. Jan.
|
||
Feb. Feb.
|
||
Mrz. Mrz.
|
||
Mär. Mär.
|
||
Apr. Apr.
|
||
Jun. Jun.
|
||
Jul. Jul.
|
||
Aug. Aug.
|
||
Sep. Sep.
|
||
Sept. Sept.
|
||
Okt. Okt.
|
||
Nov. Nov.
|
||
Dez. Dez.
|
||
Mo. Mo.
|
||
Di. Di.
|
||
Mi. Mi.
|
||
Do. Do.
|
||
Fr. Fr.
|
||
Sa. Sa.
|
||
So. So.
|
||
Std. Std.
|
||
Jh. Jh.
|
||
Jhd. Jhd.
|
||
|
||
# numbers
|
||
Tsd. Tsd.
|
||
Mio. Mio.
|
||
Mrd. Mrd.
|
||
|
||
# countries & languages
|
||
engl. engl.
|
||
frz. frz.
|
||
lat. lat.
|
||
österr. österr.
|
||
|
||
# smileys
|
||
:) :)
|
||
<3 <3
|
||
;) ;)
|
||
(: (:
|
||
:( :(
|
||
-_- -_-
|
||
=) =)
|
||
:/ :/
|
||
:> :>
|
||
;-) ;-)
|
||
:Y :Y
|
||
:P :P
|
||
:-P :-P
|
||
:3 :3
|
||
=3 =3
|
||
xD xD
|
||
^_^ ^_^
|
||
=] =]
|
||
=D =D
|
||
<333 <333
|
||
:)) :))
|
||
:0 :0
|
||
-__- -__-
|
||
xDD xDD
|
||
o_o o_o
|
||
o_O o_O
|
||
V_V V_V
|
||
=[[ =[[
|
||
<33 <33
|
||
;p ;p
|
||
;D ;D
|
||
;-p ;-p
|
||
;( ;(
|
||
:p :p
|
||
:] :]
|
||
:O :O
|
||
:-/ :-/
|
||
:-) :-)
|
||
:((( :(((
|
||
:(( :((
|
||
:') :')
|
||
(^_^) (^_^)
|
||
(= (=
|
||
o.O o.O
|
||
|
||
# single letters
|
||
a. a.
|
||
b. b.
|
||
c. c.
|
||
d. d.
|
||
e. e.
|
||
f. f.
|
||
g. g.
|
||
h. h.
|
||
i. i.
|
||
j. j.
|
||
k. k.
|
||
l. l.
|
||
m. m.
|
||
n. n.
|
||
o. o.
|
||
p. p.
|
||
q. q.
|
||
r. r.
|
||
s. s.
|
||
t. t.
|
||
u. u.
|
||
v. v.
|
||
w. w.
|
||
x. x.
|
||
y. y.
|
||
z. z.
|
||
ä. ä.
|
||
ö. ö.
|
||
ü. ü.
|