Phrase-based evaluations for the alignment shared task, as described in
Michael Carl and Sisay Fissaha, "Phrase-based Evaluation of Word-to-Word 
Alignments", in Proc. of the HLT/NAACL workshop on "Building and Using 
Parallel Texts: Data Driven Machine Translation and Beyond", Edmonton, 
Canada, May 2003. 

HIT, MISS and NOISE are the dictionary intersections of the submissions, (as
described in the paper) dic1 is the minimal dictionary, dic2 is the 
exhaustive dictionary. 
w2p.c is the evaluation, which was tested and was found to work well under 
Unix.  

Run
 
%w2p -h

to get a list of options

For any comments or questions, please contact Michael Carl, at carl@iai.uni-sb.de