ngram-entropy assumes a language model as constructed using the ngram language
modeling tools from www.opengrm.org.

A sample usage in a script is as follows:

TXT=$1
BASE=`echo $TXT | sed 's/\.txt//'`
cat $TXT | awk 'NF' >/tmp/$$
ngramsymbols --OOV_symbol="" /tmp/$$ >$BASE.syms
farcompilestrings --symbols=$BASE.syms --keep_symbols=1 /tmp/$$ >$BASE.far
ngramcount --order=2 $BASE.far >$BASE.cnts
ngrammake --method=kneser_ney $BASE.cnts >$BASE.mod
ngram-entropy $BASE.mod >$BASE.ent

To build, do the following:

1. ./configure
2. make
3. If you wish:
   make install


