Báo cáo khoa học: "Discriminative Modeling of Extraction Sets for Machine Translation" pptx

Báo cáo khoa học: "Discriminative Modeling of Extraction Sets for Machine Translation" pptx

Báo cáo khoa học: "Discriminative Modeling of Extraction Sets for Machine Translation" pptx

... Association for Computational Linguistics Discriminative Modeling of Extraction Sets for Machine Translation John DeNero and Dan Klein Computer Science Division University of California, Berkeley {denero,klein}@cs.berkeley.edu Abstract We ... sentences. Our extraction set model outperforms both un- supervised and supervised word aligners at pre- dicting word alignments and...

Ngày tải lên: 07/03/2014, 22:20

11 420 0
Báo cáo khoa học: "THE SPECIFICATION OF TIME MEANING FOR MACHINE TRANSLATION" pot

Báo cáo khoa học: "THE SPECIFICATION OF TIME MEANING FOR MACHINE TRANSLATION" pot

... syntactic/morphological forms and time meanings, resulting in a non-arbitrary relation between form and meaning. Third, we show how the calculus can be used for the interlingual analysis of the tense forms of ... relevance of the event type of the basic proposition (of. Vendler's distinction of 38 states, activities, accomplishments, achievements). Part of this ha...

Ngày tải lên: 24/03/2014, 05:21

6 359 0
Báo cáo khoa học: "Learning Non-Isomorphic Tree Mappings for Machine Translation" pptx

Báo cáo khoa học: "Learning Non-Isomorphic Tree Mappings for Machine Translation" pptx

... chart is indexed by nodes of the input tree rather than spans of the input string: 5 1. for each node c of T , in bottom-up order 2. for each q ∈ Q, let β c (q) = 0 3. for each elementary tree ... naturally extend to train on pairs of forests (including packed forests obtained by chart pars- ing). The correct tree is presumed to be an element of the forest. This makes it possi...

Ngày tải lên: 17/03/2014, 06:20

4 269 0
Báo cáo khoa học: "Discriminative Pruning of Language Models for Chinese Word Segmentation" ppt

Báo cáo khoa học: "Discriminative Pruning of Language Models for Chinese Word Segmentation" ppt

... F-Measure of 96.33%, number of bigrams decreases by up to 90%. # of bigrams % of KLD KLD 100,000 100% Step-10K 25,000 25% Step-5K 15,000 15% Step-2K 10,000 10% Table 1. Comparison of Number of ... Num(10K) F-Measure(%) KLD Discrim Step-10K Step-5K Step-2K Cut-off Figure 5. Performance Comparison of Different Pruning Methods First, we compare the performance of &qu...

Ngày tải lên: 17/03/2014, 04:20

8 294 0
Báo cáo khoa học: "A Beam-Search Extraction Algorithm for Comparable Data" pptx

Báo cáo khoa học: "A Beam-Search Extraction Algorithm for Comparable Data" pptx

... define the fertility of a source word s as the number of target words t ∈ T for which p(s|t) > ǫ. Target word cover- age and fertility are defined accordingly. A large number of ‘uncovered‘ source ... the number of extracted parallel sentence pairs significantly, which leads to a BLEU im- provement of about 1 % on our Spanish- English data. 1 Introduction The paper presents a nove...

Ngày tải lên: 17/03/2014, 02:20

4 404 0
Báo cáo khoa học: "An Evaluation of METAL: the LRC Machine Translation System" ppt

Báo cáo khoa học: "An Evaluation of METAL: the LRC Machine Translation System" ppt

... extensive use of transformations to modify structure or perform certain tests. Transformations may be used in the TEST, CONSTR, INTEr, and Transfer phases of the rules, transformations may ... optional enumeration of each constituent); a series of restrictions, which test the appropriateness of individual constituents on the right-hand side of the rule; TESTs, which enfor...

Ngày tải lên: 01/04/2014, 00:20

8 375 0
Tài liệu Báo cáo khoa học: "Fast Decoding and Optimal Decoding for Machine Translation" doc

Tài liệu Báo cáo khoa học: "Fast Decoding and Optimal Decoding for Machine Translation" doc

... fertility of e ), (the fertility of the NULL word), (the k French word produced by e in a), (the position of in f), (the position of the first fertile word to the left of e in a), (the ceiling of the ... success of any statistical machine translation system. The decoder’s job is to find the translation that is most likely according to set of previously learned parameters (and...

Ngày tải lên: 20/02/2014, 18:20

8 440 0
Báo cáo khoa học: "FUNCTIONAL UNIFICATION GRAMMAR: A FORMALISM FOR MACHINE TRANSLATION" doc

Báo cáo khoa học: "FUNCTIONAL UNIFICATION GRAMMAR: A FORMALISM FOR MACHINE TRANSLATION" doc

... parts: a formal description of the relevant facts about the language, and an interpreter of the formalism. The formal description is data whereas the interpreter is program. The formal description ... descriptions are independent of the leg of the machine they are used in and the programs are independent of the languages to which they are applied. For all the advantgages...

Ngày tải lên: 17/03/2014, 19:21

4 280 0
Báo cáo khoa học: "Disambiguating Temporal–Contrastive Discourse Connectives for Machine Translation" docx

Báo cáo khoa học: "Disambiguating Temporal–Contrastive Discourse Connectives for Machine Translation" docx

... connective word form 2. its POS tag 3. first word of argument 1 4. last word of argument 1 5. first word of argument 2 6. last word of argument 2 7. POS tag of the first word of argument 2 8. type of first ... disambiguation for translation. In addi- tion, we extracted separate training sets for each of the 8 temporal–contrastive connectives in question and one training set...

Ngày tải lên: 23/03/2014, 16:20

6 181 0
Báo cáo khoa học: "Computing Lattice BLEU Oracle Scores for Machine Translation" potx

Báo cáo khoa học: "Computing Lattice BLEU Oracle Scores for Machine Translation" potx

... q F } where the first three sets of constraints are the lin- earization of the definition of γ w , made possible by the positivity of Θ 1 and Θ 2 , and the last three sets of constraints are the path ... 26 28 30 32 34 36 Figure 2: Performance of the LB-4g oracle for differ- ent combinations of p and r on WMT11 de2en task. for matching n-grams. The values of p and r...

Ngày tải lên: 24/03/2014, 03:20

10 349 0
w