#!/bin/sh
BIN=/home/tminer/work/tmusers/wiebke/nlptools/treetagger/bin
CMD=/home/tminer/work/tmusers/wiebke/nlptools/treetagger/cmd
LIB=/home/tminer/work/tmusers/wiebke/nlptools/treetagger/lib

LEX=$LIB/chunk-lex

if [ $# -ne 2 ] ; then
	echo "ERROR: Usage: $0 <training data> <name of par file>
training-data must be of the form:
Activation-NN	NN/I-NC
of-IN	IN/B-PC
the-DT	DT/B-NC
CD28-NP	NP/I-NC
surface-NN	NN/I-NC
receptor-NN	NN/I-NC

like produced by 'cmd/tagger-chunker-english'
The resulting par file will be stored in:" $LIB
	exit 0
fi

DAT=$1
PAR=$2

echo "Lexicon generation..."
$CMD/make-chunk-lex.perl  $DAT > $LEX
$CMD/filter-corpus.perl lex $DAT > tdat

echo "Tagger training..."
$BIN/train-tree-tagger $LEX $LIB/open-class-chunks tdat $LIB/$PAR -cl 2 -dtg 0.002 -sw 70 -ecw 0.1 -atg 100.0 -st 'SENT/O'

#rm $LEX tdat