P96-1030 cases , the coverage loss due to grammar specialization was about 10 to 12 % using training
P96-1030 grammars . The coverage loss due to grammar specialization was then measured on the 1,000
E09-1093 specializa - tion . An overview of grammar specialization techniques is given in ( Sima’an
P96-1030 smaller loss of coverage due to grammar specialization . ( Recall that grammar specialization
J07-2009 particular domain , which they term grammar specialization . In effect , they induce a new
P96-1030 methods , constituent pruning and grammar specialization , are combined as follows . The
H05-2014 tools . Previously , the REGULUS grammar specialization programme has only been implemented
P96-1030 specialization . ( Recall that grammar specialization in general trades coverage for
J07-2009 data-driven , given the corpus-based grammar specialization and PCFG estimation , which the
P96-1030 method . Section 3 describes the grammar specialization method , focusing on how the
E09-1093 The cutting criteria employed in grammar specialization either require carefully manually
P96-1030 means of constituent pruning and grammar specialization based on explanation-based learning
J07-2009 various parameterizations on this grammar specialization approach -- for example , the
J12-4006 Education . <title> LFG Generation by Grammar Specialization </title> <authors></authors>
P96-1030 the coverage loss entailed by grammar specialization is reduced to approximately half
P96-1030 pruning , and is not degraded by grammar specialization . ( The slight improvement in
W00-0702 structure . <title> Corpus-Based Grammar Specialization </title> Nicola Cancedda Christer
P96-1030 in more effective pruning . 3 Grammar specialization As described in Section 1 above
P96-1030 . The loss of coverage due to grammar specialization also appears compa - rable ,
H05-2014 the grammar is very loose . The grammar specialization mechanism flattens the grammar
hide detail