Báo cáo khoa học: "Improving Tree-to-Tree Translation with Packed Forests" potx

... performance across a wide range of conditions. Suzuki et al. (2009) also experiment with the same method combined with semi-supervised learning. 699 Ciaramita and Attardi (2007) show that adding ... Pennconverter’s output are lower than with other conversions, Johansson and Nugues (2007) claim that this conversion is better suited for semantic processing, with a richer structu...

Ngày tải lên: 17/03/2014, 00:20

... Statistical Machine Translation (SMT). If it is not within the same unit, a coreferring pronoun will be trans- lated without knowledge of its antecedent, mean- ing that its translation will simply ... Machine Translation, July 5-7, Leuven, Belgium, pages 5–7. Ruslan Mitkov. 1999. Introduction: Special Issue on Anaphora Resolution in Machine Translation and Multilingual NLP. Machine Tr...

Ngày tải lên: 17/03/2014, 22:20

... detail. Training corpus To train the translation model, we translated a subset of the Europarl corpus using Its-2. The translations were then aligned with correspond- ing reference translations, resulting ... for impersonal pro-drop translation (16.78% and 29.29% respectively, Table 3). For IT→FR translation in particular, Its-2 usu- ally translates an Italian personal pro-drop with...

Ngày tải lên: 24/03/2014, 03:20

... and is p(f |e) a translation model probability. A series of increasingly sophisticated translation mod- els, referred to as the IBM Models, was defined in Brown et al. (1993). The translation model, ... quality lead to improved translation quality. 3. Experimenting with how increasing the ratio of word-aligned to sentence-aligned data affected the performance. 4. Experimenting with...

Ngày tải lên: 31/03/2014, 03:20

... a score near zero, 6 We included all translations whose probability was within a factor of 10 −4 of the highest probability translation. 7 In experiments with Urdu, which uses an Arabic-derived script, ... 1: A graphical representation of a conventional generative lexical translation model (left) and our model with an undirected translation model. For clarity, the observed node s (r...

Ngày tải lên: 17/03/2014, 00:20

... world state s, we propose a generative model p(w | s) with latent variables specifying this cor- respondence. Our model combines segmentation with align- ment. The segmentation aspect of our model ... our ex- periments, we initialized EM with a uniform dis- tribution for each multinomial and applied add-0.1 smoothing to each multinomial in the M-step. As with most complex discrete mod...

Ngày tải lên: 17/03/2014, 01:20

... probability of the translation "in order to avoid" is higher than that of the translation "can we avoid". Thus, our method selects the former as the translation. Although ... Speech Translation Evalua- tion. In Processings of the International Workshop on Spoken Language Translation 2005. Philipp Koehn, Franz J. Och, and Daniel Marcu. 2003. Statistical Phra...

Ngày tải lên: 20/02/2014, 04:20

... statistical natural language translation (SNLT). Without any predefined knowledge the system learns a statistical translation lexicon (STL), word classes (WCs) and translation rules (TRs) from ... contain words with similar syntactic/semantic properties. To arrive at WCs having both (method COMB), we determine TL WCs with the first method and afterwards we determine SL WCs...

Ngày tải lên: 08/03/2014, 05:21

... pages 558–566, Suntec, Singapore, 2-7 August 2009. c 2009 ACL and AFNLP Improving Tree-to-Tree Translation with Packed Forests Yang Liu and Yajuan L ¨ u and Qun Liu Key Laboratory of Intelligent ... de- coder runs on the translation forest. We use the cube pruning method (Chiang, 2007) to approxi- mately intersect the translation forest with the lan- guage model. Traversing the...

Ngày tải lên: 23/03/2014, 16:21

... information of sub-trees in a given parse. With help of these new features, our sys- tem achieves 93.49 F-measure with hand-crafted parses. Comparison with the best reported results, 92.0 (Xue, ... arguments of a predicate are la- beled with a contiguous sequence of integers, in the form of AN (N is a natural number); the ad- juncts are annotated as such with the label AM followed b...

Ngày tải lên: 20/02/2014, 04:20