domain adaptation of statistical parsers

Ngày tải lên: 23/03/2014, 18:20

Ngày tải lên: 17/03/2014, 00:20

Ngày tải lên: 17/03/2014, 22:20

Ngày tải lên: 23/03/2014, 16:20

Ngày tải lên: 24/03/2014, 03:20

Ngày tải lên: 30/03/2014, 21:20

... place of the out -of -domain phrase table to test the system perfor- mance. For the purpose of studying the generality of our approach, we carry out comparative experiments on two sizes of in -domain ... Effect of In -domain Monolingual Corpus Size Finally, we investigate the effect of in -domain monolingual corpus size on translation quality. In the experiment, we try different sizes of in -domain documents ... distributions of the additional in -domain monolingual corpora. As compared to the previous works, our method takes advantage of both the in -domain monolingual cor- pora and the out -of -domain bilingual...

Ngày tải lên: 19/02/2014, 19:20

... the source domain, and the real-error sentences are from the target domain. Experiments show that sta- ble improvement is achieved by using domain adaptation. 1 Introduction Case marks of a sentence ... value (real value) of the n-gram probability of the output word sequence. One feature weight is assigned. The n-gram language model can be 389 Proceedings of the 50th Annual Meeting of the Association ... classification 388 References Hal Daume, III. 2007. Frustratingly easy domain adapta- tion. In Proceedings of the 45th Annual Meeting of the Association of Computational Linguistics (ACL 2007), pages 256–263,...

Ngày tải lên: 19/02/2014, 19:20

... x. By relating the sum of the scores of all possible trees to counting the number of spanning trees in a graph, it can be shown that Z x is the determinant of the Kirchoff matrix K, which is ... marginal probability of a particular edge k → i (i.e. y i =k), the score of any edge k  → i such that k  = k is set to 0. The determinant of the resulting modi- fied Kirchoff matrix K k→i is then the sum of ... basic types of linguistic knowledge. One simple form of linguistic knowledge is the set of possible parent tags for a given child tag. This type of constraint was used in the devel- opment of a rule-based...

Ngày tải lên: 20/02/2014, 07:20

... of parent node in dependency tree. c-word: word of child node. p-pos: POS of parent node. c-pos: POS of child node. p-pos+1: POS to the right of parent in sentence. p-pos-1: POS to the left of ... parent of x j . T = {(x t , y t )} T t=1 denotes the training data. We follow the edge based factorization method of Eisner (1996) and define the score of a dependency tree as the sum of the score of ... dependency parsers that do 95 Proceedings of the 43rd Annual Meeting of the ACL, pages 91–98, Ann Arbor, June 2005. c 2005 Association for Computational Linguistics Online Large-Margin Training of Dependency...

Ngày tải lên: 20/02/2014, 15:20

... structural position of the parsing tree, the predictive information quantity of lexical feature type is larger than that of part -of- speech feature type, and the predictive information quantity of part- of- speech ... quantities of lexical feature types, part -of- speech feature types and constituent label feature types Goal One of the most important variation in statistical parsing over the last few years is that statistical lexical ... In Proceedings of the 34 th Annual Meeting of the ACL. Michael John Collins. 1997. Three generative lexicalised models for statistical parsing. In Proceedings of the 35 th Annual Meeting of the ACL. J.Eisner....

Ngày tải lên: 20/02/2014, 18:20

... rate of substitution of recombinant perlecan domain V correlated well with a comparable low rate of modi®cation of two tissue forms of perlecan [44]. An extra cysteine (C3 at position 710) in domain ... coiled-coil domains I and II of laminins [1,2], whereas o ther domains, such a s those composed of Fig. 6. Immunohistochemical localization of laminin b2 chain in basement membranes of adult mouse ... Department of Cell Biology and Physiology, Washington University School of Medicine, St Louis, MO, USA; 3 Center of Anatomy, Department of Histology, University of Go È ttingen, Germany Domain IV,...

Ngày tải lên: 21/02/2014, 03:20

... more active. Determination of E a and thermodynamic parameters for wild-type and mutant enzymes In order to elucidate the effect of mutations in terms of psychrophilic adaptation, we determined the energy of activation ... G261A/ Y269A designed on the basis of side-chain packing criteria from a modelled structure of the enzyme resulted in restor- ation of the energy of activation to the levels of the native enzyme and in ... basis of the model, the behavior of the G261A variant can be interpreted in terms of constraints introduced by the Ala side chain. The presence of the additional Gly at position 261 possibly offers...

Ngày tải lên: 22/02/2014, 04:20

... of the exonuclease and polymerase domains of Bacillus subtilis DNA polymer- ase III. Gene 111, 43–49. 15 Wieczorek A & McHenry CS (2006) The NH2-terminal php domain of the alpha subunit of ... [12,16]. The very N-ter- minal region of PolC and the C-terminal domain of DnaE appear to be specific for each type of polymer- ase. The small a ⁄ b C-terminal domain of DnaE has been shown to be responsible ... the structure of the PolC N-terminal region. We found that this region includes two consecutive structural domains. Both domains are distantly related to the structure of domain V of the clamp...

Ngày tải lên: 05/03/2014, 23:20

... which contained 0.5 lg of linearized pCOB1.8Kb and 1.0 lg of I-CsmI(217), or 1.5 lg of I-CsmI(200) or I-CsmI(237). One of the param- eters [i.e. pH, NaCl concentration, species of divalent cati- ons ... concentration of Mg 2+ was progressively detrimental to all I-CsmI polypeptides. The presence of Mg 2+ was essential for the endonuclease activity as a cofactor, while the same concentration of Mn 2+ (5 ... with 5mm of Zn 2+ ,Ca 2+ ,orCo 2+ (data not shown). Kinetic parameters of I-CsmI(200) We determined the kinetic parameters of I-CsmI(200) based on the data obtained by time course monitoring of the...

Ngày tải lên: 07/03/2014, 17:20

... the pos- sibility of domain adaptation, and the portability of the training methods of the parsers. Since the parsers listed in Section 2 have programs for the training 10 Some of the parser packages ... the domain of our target text is different from WSJ, our experiments also highlight the domain portability of parsers. We run two versions of each parser in order to investigate the two types of ... accuracy improve- ments vary when parsers are retrained with domain- specific data, indicating the importance of domain adaptation and the differences in the portability of parser training methods. Although...

Ngày tải lên: 08/03/2014, 01:20

... corpus depend on the domain from which the corpus is drawn. A change of predominant sense is often indicative of a change in domain, as different corpora drawn from different domains usually give ... @comp.nus.edu.sg Abstract Instances of a word drawn from different domains may have different sense priors (the proportions of the different senses of a word). This in turn affects the accuracy of word sense disambiguation ... empirical study of the domain dependence of supervised word sense disambiguation systems. In Proc. of EMNLP/VLC00. Adam Kilgarriff. 2001. English lexical sample task description. In Proc. of SENSEVAL-2. Yoong...

Ngày tải lên: 08/03/2014, 02:21

... change of predominant sense is often indicative of a change in domain. For example, the predominant sense of the noun interest in the BC part of the DSO corpus has the meaning “a sense of concern ... importance of the issue of domain dependence of WSD systems, but did not propose methods such as active learning or count- merging to address the specific problem of how to perform domain adaptation ... Henceforth, out -of -domain data will re- fer to BC examples, and in -domain data will refer to WSJ examples. 2.2 Choice of Nouns The WordNet Domains resource (Magnini and Cavaglia, 2000) assigns domain...

Ngày tải lên: 08/03/2014, 02:21

... great deal of at- tention. Its application to many different domains of discourse makes it an ideal candidate for domain adaptation. This work addressed two important questions of domain adaptation. ... same type of data as the polarity dataset. We also note the work of Aue and Gamon (2005), who performed a number of empirical tests on do- main adaptation of sentiment classifiers. Most of these ... train- ing set of one domain and tested on the test set of a possibly different domain. The baseline is a lin- ear classifier trained without adaptation, while the gold standard is an in -domain classifier...

Ngày tải lên: 08/03/2014, 02:21

... used for this study consist of the Wall Street Journal section of the Penn Treebank of En- glish (WSJ) and the context-free version of the NEGRA (version 2) corpus of German (Skut et al., 1997b). ... from presentation at the Institute of Cognitive Science, University of Colorado at Boulder. The au- thors are also grateful to Dan Klein and Jenny Finkel for use of maximum-entropy software they wrote. This ... type consists of a dislo- cated element coindexed with an origin site of se- mantic interpretation, as in the association in Fig- ure 1 of WHNP-1 with the direct object position of sees (a relativization),...

Ngày tải lên: 08/03/2014, 04:22