File Format

Training data

The training data should be singe TAB separated, where the first column is the word, and the last column (columns in between will be ignored) is the class of the word. Sentences are marked by a blank new line. The builtin GermaNER is based on the GermEval 2014 dataset and where only 4 named entity classes, namely ORG, PER, LOC, and OTH are used. The following is an example of how the training file looks like (tab separated token and lables, sentences separated by new line). The full training file is found here

Schartau	B-PER
sagte	O
dem	O
"	O
Tagesspiegel	B-ORG
"	O
vom	O
Freitag	O
,	O
Fischer	B-PER
sei	O
"	O
in	O
einer	O
Weise	O
aufgetreten	O
,	O
die	O
alles	O
andere	O
als	O
überzeugend	O
war	O
"	O
.	O

Firmengründer	O
Wolf	B-PER
Peter	I-PER
Bree	I-PER
arbeitete	O
Anfang	O
der	O
siebziger	O
Jahre	O
als	O
Möbelvertreter	O
,	O
als	O
er	O
einen	O
fliegenden	O
Händler	O
aus	O
dem	O
Libanon	B-LOC
traf	O
.	O

Test Data

The test data supported is a one TAB separated file, where the first column is a single word and other columns, if exist, are ignored. Sentence is marked by a blank new line.

The following example shows a two sentence test file, where the sentences are separated by a new line.

Gleich
darauf
entwirft
er
seine
Selbstdarstellung
"
Ecce
homo
"
in
enger
Auseinandersetzung
mit
diesem
Bild
Jesu

Und
in
seinen
letzten
Briefen
tritt
er
schließlich
selbst
in
diese
Erlöserrolle
ein
.