Báo cáo khoa học: "Automatic Syllabification with Structured SVMs for Letter-To-Phoneme Conversion" doc

... report L2P performance without any syllabification infor- mation, with perfect dictionary syllabification, and with our small learned models of syllabification. L2P performance with dictionary syllabification ... (Tsochantaridis et al., 2004). 4 Syllabification with Structured SVMs In this paper we apply structured SVMs to the syl- labification problem. Specifically, we form...

Ngày tải lên: 17/03/2014, 02:20

... empirically that the structured preferences are crucial to the suc- cess of our task. Our model, trained with- out annotated data and with a small number of structured preferences, yields performance competitive ... TIME-WITHIN” is in general considered a better sub-structure than “said TIME-WITHIN” for the “Attack” event, we may want to assign structured preference to a complete...

Ngày tải lên: 30/03/2014, 17:20

... DNA modification with cisplatin [31,33,34]. The ability of p53 to recognize the cisPt-DNA was more pronounced for the post-translationally unmodified (‘latent’) form of the protein than for its ‘activated’ forms ... of hydrogen bond- ing within the base pairs [26–28]. Cisplatin adducts occurring within p53DBS can therefore be expected to cause severe deformations of the binding site with...

Ngày tải lên: 19/02/2014, 05:20

... 53 72 68 69 Table 4: The performance of IB and ID for the 3 levels of class hierarchy for informative values of K pairwise similarities, NN shows mostly better per- formance than IB and ID on ... only 4 . 3.2.3 Information Bottleneck The Information Bottleneck (Tishby et al., 1999) (IB) is an information-theoretic method which controls the balance between: (i) the loss of information by r...

Ngày tải lên: 31/03/2014, 01:20

... were screened for repression in the presence of 0.4 lm 4-ddma-atc and rescreened for induction with 0.4 lm atc and without inducer on MacConkey agar plates. We screened 17 600 colonies with mutations ... binding to tetO without effector and with atc. The strongest tetO binding is observed for TetR i2 -V99N with 4-ddma-atc as corepressor, while TetR i2 -L17A does not show a clear...

Ngày tải lên: 20/02/2014, 01:20

... developed a super- vised classifier for identifying speculation cues and a manually compiled list of lexico-syntactic rules for identifying their scopes. For the performance of the rule based system ... patterns automatically extracted from a corpus anno- tated with negation/speculation cues and their scopes (the BioScope corpus). The system performs on par with state-of-the-art mach...

Ngày tải lên: 20/02/2014, 04:20

... transducers that accept underlying forms as input and generate surface forms as output. The algorithm for learning them is an extension of the OSTIA algorithm for learn- ing general subsequential ... the error performance on test data. The improved algorithm induced a flapping transducer with the minimum number of states with as few as 6250 samples. The use of alignment informat...

Ngày tải lên: 08/03/2014, 07:20

... MT results were improved by the latter method. 1 Introduction Along with the efforts made to accumulate bilin- gual corpora for many language pairs, quite a few machine translation (MT) systems ... measures are assumed to be available for controlled translation. First three measures are for each of the bilingual sentences in the corpus and the fourth measure is for the whole corpus:...

Ngày tải lên: 08/03/2014, 21:20

... part-of-speech string associated with words with multivowel string kernels will be NA. Rule B: The part-of-speech string associated with a word ending in ed will be PP, and with a word end- ing in ... Reference 1. In its modi- fied form, as stated in Rule A and tested by our five- hundred-word sample, it is accurate for only about 55-60 per cent of the cases, but is good fo...

Ngày tải lên: 16/03/2014, 19:20

... Association for Computational Linguistics, pages 1365–1374, Portland, Oregon, June 19-24, 2011. c 2011 Association for Computational Linguistics Discovering Sociolinguistic Associations with Structured ... and implemen- tation for output-regression with structured sparsity; in Section 4 and 5 we give the details of its applica- tion to select terms and demographic features. Fo...

Ngày tải lên: 17/03/2014, 00:20