P11-2035 opportunity for large-scale , robust parser combination . 5 Conclusion We have presented
P09-2010 locate the pre cise benefits of the parser combination . We will also investigate the
P07-1079 accuracy simply by using an existing parser combination approach . As a first step ,
N09-2066 new linear algorithm to perform parser combination . Experiments on the corpora
N09-2064 improve upon existing methods for parser combination . First , we propose a method
P07-1079 experiments on development data using parser combinations to produce dependency constraints
D09-1161 confidence " was used in prior parser combination studies to refer to the accuracy
N09-2064 improve upon existing methods for parser combination . First , while constituent recombination
N09-2064 framework . Third , we extend these parser combination methods from 1-best outputs to
N09-2064 selection . Third , we extend these parser combination methods from multiple 1-best
N06-2033 <authors></authors> Abstract We present a novel parser combination scheme that works by reparsing
D08-1017 reinterpreting and generalizing their parser combination scheme as a stacking of parsers
D12-1105 in effect , a simple method for parser combination , not all that dissimilar to
N09-2066 are obtained by using a linear parser combination algorithm on the outputs of three
P07-1079 in previous work on dependency parser combination ( Zeman and &#711;Zabokrtsk &#180;
N09-2066 2007 Shared Task , the linear parser combination achieves the best LAS for Czech
N10-1091 Experiments 3.1 On scoring models for parser combination The most common approach for
P10-1110 learning ( Koo et al. , 2008 ) and parser combination ( Zhang and Clark , 2008 ) do
N06-2033 No . HR0011-06-C-0023 . <title> Parser Combination by Reparsing Sagae Language Technologies
N10-1091 learning time , we argue that runtime parser combination is a more attractive approach
hide detail