Báo cáo khoa học: "Consistent Translation using Discriminative Learning: A Translation Memory-inspired Approach" pdf

... phrase-based SMT, translation models and lan- guage models are automatically learned and/or gen- eralised from the training data, and a translation is produced by maximising a weighted combination ... decide to mark up the input sentence if the translations of the marked phrases are accurate when taken contextual information into account. As large-scale manually annotated data is not...

Ngày tải lên: 07/03/2014, 22:20

... “The resignation of Khaddam was abrupt” as an example. In particular, the “depth” features at- tempt to capture the “importance” the dependency match, as measured by the depth of the ancestor in the ... instances that had been marked relevant or irrelevant. We trained the kernel version of an averaged per- ceptron model (Freund and Schapire, 1999), using a polynomial kernel with degree 4...

Ngày tải lên: 20/02/2014, 09:20

... War and Peace.”), and taking such variations into account can improve system performance significantly (Harabagiu et al., 2003; Harabagiu and Hickl, 2006). A paraphrase gener- ator, meaning a ... Finally, Pang et al. (2003) create finite state automata by merging parse trees of aligned sentences from a parallel corpus; in each automaton, different paths represent paraphrases. Again, a p...

Ngày tải lên: 08/03/2014, 01:20

... setiap satu yang boleh menampung kelas seramai 30 pelajar, selain bekalan-bekalan lain seperti 500 khemah biasa, barang makanan dan ubat-ubatan untuk mangsa gempa Sichuan. ref1: Mercy Relief has ... regarded as separate languages, which are mutually intelligible, but occasionally dif- fer in orthography/pronunciation and vocabulary: Bahasa Malaysia (lit. ‘language of Malaysia’) and Bahasa ......

Ngày tải lên: 07/03/2014, 22:20

... less than 1% after all the available WSJ adaptation examples are added as additional training data. To obtain a clearer picture of the adaptation process, we discard these 4 nouns, leaving a set ... in BC and WSJ, average MFS accuracy, average number of BC training, and WSJ adaptation examples per noun. data, and the rest of the WSJ examples are desig- nated as in-domain adaptation data....

Ngày tải lên: 08/03/2014, 02:21

... given source language text. The target language speech is a human-produced translation of the source language text. Statistical machine translation (MT) models are employed to take into account the ... be represented as a large word graph. We consider MT and ASR word graphs as FSA. Then, we are able to use FSA algorithms to integrate MT and ASR word graphs. The FSA implementation of t...

Ngày tải lên: 20/02/2014, 12:20

... undergoing phase II/III human clinical trials as an antitumor drug. Native onconase from Rana pipiens and its amphibian homologs have an N-ter- minal pyroglutamyl residue that is essential for obtaining fully ... calculate the fraction of each species present at a particular instant and under particular conditions. This strategy was used to characterize the factors that affect the reactions...

Ngày tải lên: 19/02/2014, 12:20

... similarities are hierar- chical rules. 2 Hierarchical phrase-based MT system The hierarchical phrase-based translation method (Chiang, 2005; Chiang, 2007) is a formal syntax- based translation ... addi- tional features of the translation model to im- prove translation performance. Significant im- provements are obtained over a state-of-the-art hierarchical phrase-based machine...

Ngày tải lên: 20/02/2014, 04:20

... h: • a partial translation x; • a coverage set of input words c; • a translation score σ. In the case of non-projective dependency parsing, we need to maintain additional information for each ... set of all tags appearing between each pair of words. Ex- tracting all these tags takes time O(n) for any arbi- trary pair (i, j). Since i and j are both free vari- ables, feature computati...

Ngày tải lên: 20/02/2014, 07:20

... coherence in a manner that takes advantage of their individual strengths. An integral part of this framework are algo- rithms for searching and training these stochastic coherence models. We evaluate the ... results. 4.1 Evaluation setting The task on which we conduct our evaluation is information ordering (Lapata, 2003; Barzilay and Lee, 2004; Barzilay and Lapata, 2005). In this task, a...

Ngày tải lên: 20/02/2014, 12:20