| Score1 |
Output |
Command |
Status |
Score2 |
Error |
|
| 0.2837 |
MT evaluation scorer began on 2008 Feb 19 at 16:28:25
command line: /disk4/eval-site/tools/bleu/mteval-v11b.pl -s /disk4/eval-site/data/europarl-v3/test/test/test2007-src.de.sgm -r /disk4/eval-site/data/europarl-v3/test/test/test2007-ref.en.sgm -t /disk4/html/eval/public/data/U-3_S-1_T-1_test2007.detokenized.sgm.35
Evaluation of any-to-en translation using:
src set "test2007" (1 docs, 2000 segs)
ref set "test2007" (1 refs)
tst set "test2007" (1 systems)
length ratio: 0.997034752698375 (58842/59017), penalty (log): -0.00297406614323092
NIST score = 7.1756 BLEU score = 0.2837 for system "Edinburgh"
# ------------------------------------------------------------------------
Individual N-gram scoring
1-gram 2-gram 3-gram 4-gram 5-gram 6-gram 7-gram 8-gram 9-gram
------ ------ ------ ------ ------ ------ ------ ------ ------
NIST: 5.1088 1.5603 0.3969 0.0844 0.0252 0.0088 0.0042 0.0025 0.0007 "Edinburgh"
BLEU: 0.6230 0.3445 0.2157 0.1416 0.0965 0.0679 0.0494 0.0365 0.0272 "Edinburgh"
# ------------------------------------------------------------------------
Cumulative N-gram scoring
1-gram 2-gram 3-gram 4-gram 5-gram 6-gram 7-gram 8-gram 9-gram
------ ------ ------ ------ ------ ------ ------ ------ ------
NIST: 5.1088 6.6691 7.0660 7.1504 7.1756 7.1844 7.1886 7.1911 7.1918 "Edinburgh"
BLEU: 0.6212 0.4619 0.3580 0.2837 0.2285 0.1866 0.1542 0.1288 0.1083 "Edinburgh"
MT evaluation scorer ended on 2008 Feb 19 at 16:29:28
|
/disk4/eval-site/tools/bleu/mteval-v11b.pl -s /disk4/eval-site/data/europarl-v3/test/test/test2007-src.de.sgm -r /disk4/eval-site/data/europarl-v3/test/test/test2007-ref.en.sgm -t /disk4/html/eval/public/data/U-3_S-1_T-1_test2007.detokenized.sgm.35 |
2 |
0.997035 |
|
Show |
| 0.644588 |
"/disk4/html/eval/public/data/U-3_S-1_T-1_test2007.detokenized.sgm.35" was successfully parsed as SGML
"/disk4/eval-site/data/europarl-v3/test/test/test2007-ref.en.sgm" was successfully parsed as SGML
Total TER: 0.6445876529373307 (34508.0/53535.0)
Number of calls to beam search: 49003
Number of segments scored: 2000
Number of shifts tried: 47003
|
/usr/java/jdk1.5.0_14/bin/java -jar /disk4/eval-site/tools/ter/tercom.jar -r /disk4/eval-site/data/europarl-v3/test/test/test2007-ref.en.sgm -h /disk4/html/eval/public/data/U-3_S-1_T-1_test2007.detokenized.sgm.35 |
2 |
|
|
Show |
| 0.1515 |
MT evaluation scorer began on 2008 Feb 23 at 17:11:52
command line: /disk4/eval-site/tools/bleu/mteval-v11b.pl -s /disk4/eval-site/data/europarl-v3/test/devtest/devtest2006-src.el.sgm -r /disk4/eval-site/data/europarl-v3/test/devtest/devtest2006-ref.fi.sgm -t /disk4/html/eval/public/data/20080223170310_U-3_S-8_T-378_devtest2006.detokenized.sgm.100
Evaluation of any-to-fi translation using:
src set "devtest2006" (1 docs, 2000 segs)
ref set "devtest2006" (1 refs)
tst set "devtest2006" (1 systems)
length ratio: 1.00283899528438 (41682/41564), penalty (log): 0
NIST score = 4.8117 BLEU score = 0.1515 for system "Edinburgh"
# ------------------------------------------------------------------------
Individual N-gram scoring
1-gram 2-gram 3-gram 4-gram 5-gram 6-gram 7-gram 8-gram 9-gram
------ ------ ------ ------ ------ ------ ------ ------ ------
NIST: 3.9558 0.6803 0.1328 0.0336 0.0092 0.0024 0.0011 0.0006 0.0002 "Edinburgh"
BLEU: 0.4554 0.1932 0.1022 0.0586 0.0359 0.0234 0.0159 0.0113 0.0084 "Edinburgh"
# ------------------------------------------------------------------------
Cumulative N-gram scoring
1-gram 2-gram 3-gram 4-gram 5-gram 6-gram 7-gram 8-gram 9-gram
------ ------ ------ ------ ------ ------ ------ ------ ------
NIST: 3.9558 4.6361 4.7689 4.8025 4.8117 4.8140 4.8152 4.8157 4.8160 "Edinburgh"
BLEU: 0.4554 0.2966 0.2080 0.1515 0.1136 0.0873 0.0684 0.0546 0.0443 "Edinburgh"
MT evaluation scorer ended on 2008 Feb 23 at 17:12:46
|
/disk4/eval-site/tools/bleu/mteval-v11b.pl -s /disk4/eval-site/data/europarl-v3/test/devtest/devtest2006-src.el.sgm -r /disk4/eval-site/data/europarl-v3/test/devtest/devtest2006-ref.fi.sgm -t /disk4/html/eval/public/data/20080223170310_U-3_S-8_T-378_devtest2006.detokenized.sgm.100 |
2 |
1.00284 |
|
Show |
| 0.804755 |
"/disk4/html/eval/public/data/20080223170310_U-3_S-8_T-378_devtest2006.detokenized.sgm.100" was successfully parsed as SGML
"/disk4/eval-site/data/europarl-v3/test/devtest/devtest2006-ref.fi.sgm" was successfully parsed as SGML
Total TER: 0.8047546359113523 (28469.0/35376.0)
Number of calls to beam search: 11945
Number of segments scored: 2000
Number of shifts tried: 9945
|
/usr/java/jdk1.5.0_14/bin/java -jar /disk4/eval-site/tools/ter/tercom.jar -r /disk4/eval-site/data/europarl-v3/test/devtest/devtest2006-ref.fi.sgm -h /disk4/html/eval/public/data/20080223170310_U-3_S-8_T-378_devtest2006.detokenized.sgm.100 |
2 |
|
|
Show |
| 0.1481 |
MT evaluation scorer began on 2008 Feb 23 at 17:12:23
command line: /disk4/eval-site/tools/bleu/mteval-v11b.pl -s /disk4/eval-site/data/europarl-v3/test/devtest/test2006-src.el.sgm -r /disk4/eval-site/data/europarl-v3/test/devtest/test2006-ref.fi.sgm -t /disk4/html/eval/public/data/20080223170310_U-3_S-8_T-268_test2006.detokenized.sgm.100
Evaluation of any-to-fi translation using:
src set "test2006" (1 docs, 2000 segs)
ref set "test2006" (1 refs)
tst set "test2006" (1 systems)
length ratio: 1.00096550100082 (42506/42465), penalty (log): 0
NIST score = 4.8085 BLEU score = 0.1481 for system "Edinburgh"
# ------------------------------------------------------------------------
Individual N-gram scoring
1-gram 2-gram 3-gram 4-gram 5-gram 6-gram 7-gram 8-gram 9-gram
------ ------ ------ ------ ------ ------ ------ ------ ------
NIST: 3.9555 0.6836 0.1288 0.0321 0.0084 0.0027 0.0010 0.0004 0.0001 "Edinburgh"
BLEU: 0.4551 0.1930 0.0992 0.0552 0.0334 0.0212 0.0140 0.0095 0.0066 "Edinburgh"
# ------------------------------------------------------------------------
Cumulative N-gram scoring
1-gram 2-gram 3-gram 4-gram 5-gram 6-gram 7-gram 8-gram 9-gram
------ ------ ------ ------ ------ ------ ------ ------ ------
NIST: 3.9555 4.6391 4.7679 4.8000 4.8085 4.8112 4.8122 4.8126 4.8126 "Edinburgh"
BLEU: 0.4551 0.2964 0.2058 0.1481 0.1100 0.0836 0.0647 0.0509 0.0406 "Edinburgh"
MT evaluation scorer ended on 2008 Feb 23 at 17:13:20
|
/disk4/eval-site/tools/bleu/mteval-v11b.pl -s /disk4/eval-site/data/europarl-v3/test/devtest/test2006-src.el.sgm -r /disk4/eval-site/data/europarl-v3/test/devtest/test2006-ref.fi.sgm -t /disk4/html/eval/public/data/20080223170310_U-3_S-8_T-268_test2006.detokenized.sgm.100 |
2 |
1.00097 |
|
Show |
| 0.812422 |
"/disk4/html/eval/public/data/20080223170310_U-3_S-8_T-268_test2006.detokenized.sgm.100" was successfully parsed as SGML
"/disk4/eval-site/data/europarl-v3/test/devtest/test2006-ref.fi.sgm" was successfully parsed as SGML
Total TER: 0.8124222228367578 (29378.0/36161.0)
Number of calls to beam search: 13150
Number of segments scored: 2000
Number of shifts tried: 11150
|
/usr/java/jdk1.5.0_14/bin/java -jar /disk4/eval-site/tools/ter/tercom.jar -r /disk4/eval-site/data/europarl-v3/test/devtest/test2006-ref.fi.sgm -h /disk4/html/eval/public/data/20080223170310_U-3_S-8_T-268_test2006.detokenized.sgm.100 |
2 |
|
|
Show |
| 0.1438 |
MT evaluation scorer began on 2008 Feb 23 at 17:12:58
command line: /disk4/eval-site/tools/bleu/mteval-v11b.pl -s /disk4/eval-site/data/europarl-v3/test/test/test2007-src.el.sgm -r /disk4/eval-site/data/europarl-v3/test/test/test2007-ref.fi.sgm -t /disk4/html/eval/public/data/20080223170310_U-3_S-8_T-28_test2007.detokenized.sgm.100
Evaluation of any-to-fi translation using:
src set "test2007" (1 docs, 2000 segs)
ref set "test2007" (1 refs)
tst set "test2007" (1 systems)
length ratio: 0.995672121842777 (42101/42284), penalty (log): -0.00434669010237276
NIST score = 4.7404 BLEU score = 0.1438 for system "Edinburgh"
# ------------------------------------------------------------------------
Individual N-gram scoring
1-gram 2-gram 3-gram 4-gram 5-gram 6-gram 7-gram 8-gram 9-gram
------ ------ ------ ------ ------ ------ ------ ------ ------
NIST: 3.9156 0.6645 0.1230 0.0295 0.0078 0.0029 0.0015 0.0010 0.0002 "Edinburgh"
BLEU: 0.4503 0.1892 0.0957 0.0534 0.0323 0.0206 0.0135 0.0093 0.0068 "Edinburgh"
# ------------------------------------------------------------------------
Cumulative N-gram scoring
1-gram 2-gram 3-gram 4-gram 5-gram 6-gram 7-gram 8-gram 9-gram
------ ------ ------ ------ ------ ------ ------ ------ ------
NIST: 3.9156 4.5801 4.7031 4.7326 4.7404 4.7433 4.7449 4.7459 4.7461 "Edinburgh"
BLEU: 0.4483 0.2906 0.2004 0.1438 0.1066 0.0810 0.0627 0.0493 0.0396 "Edinburgh"
MT evaluation scorer ended on 2008 Feb 23 at 17:13:54
|
/disk4/eval-site/tools/bleu/mteval-v11b.pl -s /disk4/eval-site/data/europarl-v3/test/test/test2007-src.el.sgm -r /disk4/eval-site/data/europarl-v3/test/test/test2007-ref.fi.sgm -t /disk4/html/eval/public/data/20080223170310_U-3_S-8_T-28_test2007.detokenized.sgm.100 |
2 |
0.995672 |
|
Show |
| 0.811326 |
"/disk4/html/eval/public/data/20080223170310_U-3_S-8_T-28_test2007.detokenized.sgm.100" was successfully parsed as SGML
"/disk4/eval-site/data/europarl-v3/test/test/test2007-ref.fi.sgm" was successfully parsed as SGML
Total TER: 0.8113259821576994 (29284.0/36094.0)
Number of calls to beam search: 12231
Number of segments scored: 2000
Number of shifts tried: 10231
|
/usr/java/jdk1.5.0_14/bin/java -jar /disk4/eval-site/tools/ter/tercom.jar -r /disk4/eval-site/data/europarl-v3/test/test/test2007-ref.fi.sgm -h /disk4/html/eval/public/data/20080223170310_U-3_S-8_T-28_test2007.detokenized.sgm.100 |
2 |
|
|
Show |
| 0.1881 |
MT evaluation scorer began on 2008 Feb 23 at 17:13:29
command line: /disk4/eval-site/tools/bleu/mteval-v11b.pl -s /disk4/eval-site/data/europarl-v3/test/devtest/devtest2006-src.en.sgm -r /disk4/eval-site/data/europarl-v3/test/devtest/devtest2006-ref.el.sgm -t /disk4/html/eval/public/data/20080223170310_U-3_S-9_T-386_devtest2006.detokenized.sgm.101
Evaluation of any-to-el translation using:
src set "devtest2006" (1 docs, 2000 segs)
ref set "devtest2006" (1 refs)
tst set "devtest2006" (1 systems)
length ratio: 0.99665012189347 (56826/57017), penalty (log): -0.00336113750747891
NIST score = 5.4104 BLEU score = 0.1881 for system "Edinburgh"
# ------------------------------------------------------------------------
Individual N-gram scoring
1-gram 2-gram 3-gram 4-gram 5-gram 6-gram 7-gram 8-gram 9-gram
------ ------ ------ ------ ------ ------ ------ ------ ------
NIST: 4.1959 0.9760 0.1944 0.0362 0.0079 0.0024 0.0010 0.0002 0.0001 "Edinburgh"
BLEU: 0.4914 0.2335 0.1348 0.0821 0.0513 0.0332 0.0217 0.0148 0.0104 "Edinburgh"
# ------------------------------------------------------------------------
Cumulative N-gram scoring
1-gram 2-gram 3-gram 4-gram 5-gram 6-gram 7-gram 8-gram 9-gram
------ ------ ------ ------ ------ ------ ------ ------ ------
NIST: 4.1959 5.1719 5.3663 5.4025 5.4104 5.4128 5.4138 5.4140 5.4141 "Edinburgh"
BLEU: 0.4898 0.3376 0.2483 0.1881 0.1450 0.1134 0.0895 0.0714 0.0576 "Edinburgh"
MT evaluation scorer ended on 2008 Feb 23 at 17:14:49
|
/disk4/eval-site/tools/bleu/mteval-v11b.pl -s /disk4/eval-site/data/europarl-v3/test/devtest/devtest2006-src.en.sgm -r /disk4/eval-site/data/europarl-v3/test/devtest/devtest2006-ref.el.sgm -t /disk4/html/eval/public/data/20080223170310_U-3_S-9_T-386_devtest2006.detokenized.sgm.101 |
2 |
0.99665 |
|
Show |
| 0.527945 |
"/disk4/html/eval/public/data/20080223170310_U-3_S-9_T-386_devtest2006.detokenized.sgm.101" was successfully parsed as SGML
"/disk4/eval-site/data/europarl-v3/test/devtest/devtest2006-ref.el.sgm" was successfully parsed as SGML
Total TER: 0.5279444824552829 (27007.0/51155.0)
Number of calls to beam search: 229146
Number of segments scored: 2000
Number of shifts tried: 227146
|
/usr/java/jdk1.5.0_14/bin/java -jar /disk4/eval-site/tools/ter/tercom.jar -r /disk4/eval-site/data/europarl-v3/test/devtest/devtest2006-ref.el.sgm -h /disk4/html/eval/public/data/20080223170310_U-3_S-9_T-386_devtest2006.detokenized.sgm.101 |
2 |
|
|
Show |