Báo cáo khoa học: "Minimum Error Rate Training in Statistical Machine Translation" potx

... obtained by simply summing over scores for individual sentences. 4 Training Criteria for Minimum Error Rate Training In the following, we assume that we can measure the number of errors in sentence ... Minimum Error Rate Training in Statistical Machine Translation Franz Josef Och Information Sciences Institute University of Southern California 4676 Admiralty Way, Suite...

Ngày tải lên: 23/03/2014, 19:20

... domains: text news, trained on a large cor- pus, and spoken travel conversation, trained on a sig- nificantly smaller corpus. We show that segmenting the Arabic target in training and decoding ... 200,000 word training set, a 500 sentence tuning set and a 500 sentence test set. We use the Arabic side of the training data to train the language model and use trigrams for the baseline sys...

Ngày tải lên: 20/02/2014, 09:20

... change in perfor- mance between training on the original training data in Eq. 2 or on the modified training data in Eq. 10. Line shows that even when training the float weights on an event set obtained ... results in line - are obtained by training ’float’ weights only. Here, the training is carried out by running only once over % of the training data. The model including t...

Ngày tải lên: 20/02/2014, 15:20

... shown in Table 5. 6.2 Training and test perplexities In order to compute the training and test perplex- ities, we split the whole aligned training corpus in two parts as shown in Table 6. The training and ... computed automatically using another statistical train- ing procedure (Och, 1999) which often pro- duces word classes including words with the same semantic meaning in...

Ngày tải lên: 20/02/2014, 18:20

... de- tail. Finally, experimental results for a bilingual cor- pus are reported. 1.1 Statistical Machine Translation In statistical machine translation, the goal of the search strategy can ... spoken dialogs in the domain of appointment sche- duling (Wahlster, 1993). German source sentences are translated into English. In Table 1 the character- istics of the training and...

Ngày tải lên: 20/02/2014, 18:20

... weakness in order model- ing, without affecting its strengths. To fur- ther increase flexibility, we incorporate cohe- sion as a decoder feature, creating a soft con- straint. The resulting cohesive, ... according to Minimum Error Rate Training or MERT (Och, 2003). Phrasal SMT draws strength from being able to memorize non- compositional and context-specific translations, as well as loc...

Ngày tải lên: 08/03/2014, 01:20

... data, train the language models and the translation models. It also contains tools for tuning these models using minimum error rate training (Och 2003) and evalu- ating the resulting translations ... additional sources of informa- tion have been shown to be valuable when inte- grated into pre-processing or post-processing steps. Moses also integrates confusion network de- coding, wh...

Ngày tải lên: 23/03/2014, 18:20

... probabilities increases the robustness. Wrong phrase pairs resulting from errors in the Viterbi alignment will have a low probability. 3 What's in the Training Data 3.1 The Corpora To train the Chinese-to-English ... covered by the training data. Coverage can be expressed in terms of tokens, i.e. how many of the tokens in the test sentences are covered by the vocabulary of th...

Ngày tải lên: 24/03/2014, 03:20

... re- laxing the clipping constraints: starting from an unconstrained problem, the counts clipping is en- forced by incrementally strengthening the weight of paths satisfying the constraints. The ... defined by Equation (4). In the beginning each arc is initialized with a singleton set containing one tuple with a single word as the partial hypothesis. For the semiring operations we define one ....

Ngày tải lên: 24/03/2014, 03:20

... answering questions, making suggestions and providing sup- port. References Birch, A., Osborne, M., and Koehn, P. 2007. CCG Supertags in factored Statistical Machine Translation. In Proceedings ... Computa- tional Linguistics. The Association for Computer Lin- guistics. Carpuat, M. and Wu, D. 2007. Improving Statistical Ma- chine Translation using Word Sense Disambiguation. In Pro...

Ngày tải lên: 31/03/2014, 00:20