Zobrazeno 1 - 4
of 4
pro vyhledávání: '"Markos Mylonakis"'
Publikováno v:
Journal of Logic and Computation, 24(2), 433-453. Oxford University Press
This article uses semi-supervised Expectation Maximization (EM) to learn lexico-syntactic dependencies, i.e. associations between words and the structures that occur with them. Due to Zipfian distributions in language, such dependencies are extremely
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::8bddbd26b1efef648753df6dc962c8ec
https://doi.org/10.1093/logcom/exs062
https://doi.org/10.1093/logcom/exs062
Autor:
Khalil Sima'an, Markos Mylonakis
Publikováno v:
EMNLP 2008: 2008 Conference on Empirical Methods in Natural Language Processing: Proceedings of the conference, 630-639
STARTPAGE=630;ENDPAGE=639;TITLE=EMNLP 2008: 2008 Conference on Empirical Methods in Natural Language Processing: Proceedings of the conference
EMNLP
STARTPAGE=630;ENDPAGE=639;TITLE=EMNLP 2008: 2008 Conference on Empirical Methods in Natural Language Processing: Proceedings of the conference
EMNLP
The conditional phrase translation probabilities constitute the principal components of phrase-based machine translation systems. These probabilities are estimated using a heuristic method that does not seem to optimize any reasonable objective funct
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::94e9d617730ba36ca7e8aadd3e8e570b
https://dare.uva.nl/personal/pure/en/publications/phrase-translation-probabilities-with-itg-priors-and-smoothing-as-learning-objective(b2c08b75-9ffb-4217-89cc-f62f53e233e3).html
https://dare.uva.nl/personal/pure/en/publications/phrase-translation-probabilities-with-itg-priors-and-smoothing-as-learning-objective(b2c08b75-9ffb-4217-89cc-f62f53e233e3).html
Autor:
Markos Mylonakis, Khalil Sima'an
Publikováno v:
SLT 2008: 2008 IEEE Workshop on Spoken Language Technology: Proceedings, 237-240
STARTPAGE=237;ENDPAGE=240;TITLE=SLT 2008: 2008 IEEE Workshop on Spoken Language Technology: Proceedings
SLT
STARTPAGE=237;ENDPAGE=240;TITLE=SLT 2008: 2008 IEEE Workshop on Spoken Language Technology: Proceedings
SLT
The heuristic estimates of conditional phrase translation probabilities are based on frequency counts in a word-aligned parallel corpus. Earlier attempts at more principled estimation using Expectation-Maximization (EM) under perform this heuristic.
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::b998964b0ecf8c1d626e3f6d91909c32
https://dare.uva.nl/personal/pure/en/publications/better-statistical-estimation-can-benefit-all-phrases-in-phrasebased-statistical-machine-translation(8c3e4c58-2b01-461a-b591-22e1d1fd9c50).html
https://dare.uva.nl/personal/pure/en/publications/better-statistical-estimation-can-benefit-all-phrases-in-phrasebased-statistical-machine-translation(8c3e4c58-2b01-461a-b591-22e1d1fd9c50).html
Publikováno v:
ICML
ACM International Conference Proceedings Series, 227, 665-672
ACM International Conference Proceedings Series, 227, 665-672
Shannon's Noisy-Channel model, which describes how a corrupted message might be reconstructed, has been the corner stone for much work in statistical language and speech processing. The model factors into two components: a language model to character
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::30b954bb4398a75b405e42d07eca8389
https://doi.org/10.1145/1273496.1273580
https://doi.org/10.1145/1273496.1273580