moses-smt/mosesdecoder
1
1
Fork 0
mirror of https://github.com/moses-smt/mosesdecoder.git synced 2024-12-28 14:32:38 +03:00
Code Issues Packages Projects Releases Wiki Activity
a300824bd1
mosesdecoder/scripts/analysis
History
Ondrej Bojar 01e364d1e6 use --n=0 to check coverage of full sents
2014-09-10 14:33:55 +02:00
..
perllib fix start weights in experiment.perl, add hypothesis queue for picking hope and fear translations, add variations to 1slack formulation 2012-06-01 01:49:42 +01:00
smtgui revert mode changes 2012-07-04 12:25:21 +01:00
bootstrap-hypothesis-difference-significance.pl added support for arbitrary encodings via the $IO_ENCODING global variable on line 23; set to UTF8 by default 2010-11-29 09:04:44 +00:00
extract-target-trees.py extract-target-trees.py: update to match current trace format 2013-11-10 16:09:06 +00:00
nontranslated_words.pl add svn id comments to start of file 2007-03-14 22:22:36 +00:00
oov.pl use --n=0 to check coverage of full sents 2014-09-10 14:33:55 +02:00
README fix start weights in experiment.perl, add hypothesis queue for picking hope and fear translations, add variations to 1slack formulation 2012-06-01 01:49:42 +01:00
sentence-by-sentence.pl create valid html header, according to Tomas Hudik 2011-10-12 10:18:36 +00:00
sg2dot.perl fixed regexes to read current -osg format 2010-02-03 14:35:21 +00:00
show-phrases-used.pl just setting the executable bit 2010-01-29 19:49:37 +00:00
suspicious_tokenization.pl list frequent mismatched tokenizations first 2010-02-03 16:37:08 +00:00
weight-scan-summarize.sh Ondrej's little tools to examine weight settings 2011-07-08 00:11:10 +00:00
weight-scan.pl Change Bin to RealBin. Thanks to Tom Hoar 2012-06-26 11:57:23 -04:00

README 

Put any scripts useful for human analysis of MT output here.

sentence-by-sentence.pl [EVH]: show comparison of sentences in reference translation(s)/system output(s)/(truth) in colorful format
-- show all sentences given, with non-matching words in the system output marked, BLEU scores given by sentence, and matching n-grams shown in a table
-- requires all input files be utf8-encoded (you can convert a file with `cat FILE | perl -n -e 'binmode(STDOUT, ":utf8"); print;' > FILE.utf8`)

show-phrases-used.pl [EVH]: draw colorful diagram of which source phrases map to which target phrases
-- requires the Perl GD module, which in turn requires that gd be installed and in LD_LIBRARY_PATH
-- show average length of source phrases used for each sentence and overall
-- command-line options -r for reference and -s for source; lone filenames are taken to be system outputs