Lexicalized Stochastic Modeling of Constraint-Based Grammars using Log-Linear Measures and EM Training

Simple item page
dc.creatorRiezler, Stefan
dc.creatorPrescher, Detlef
dc.creatorKuhn, Jonas
dc.creatorJohnson, Mark
dc.date2000-08-30
dc.date.accessioned2026-07-07T03:16:31Z
dc.date.available2026-07-07T03:16:31Z
dc.descriptionWe present a new approach to stochastic modeling of constraint-based grammars that is based on log-linear models and uses EM for estimation from unannotated data. The techniques are applied to an LFG grammar for German. Evaluation on an exact match task yields 86% precision for an ambiguity rate of 5.4, and 90% precision on a subcat frame match for an ambiguity rate of 25. Experimental comparison to training from a parsebank shows a 10% gain from EM training. Also, a new class-based grammar lexicalization is presented, showing a 10% gain over unlexicalized models.
dc.description8 pages, uses acl2000.sty
dc.identifierhttps://arxiv.org/abs/cs/0008034
dc.identifierhttp://arxiv.org/abs/cs/0008034
dc.identifierProceedings of the 38th Annual Meeting of the ACL, 2000
dc.identifier.urihttp://salesiana.dossiersoluciones.com/handle/123456789/30381
dc.subjectComputation and Language
dc.subjectI.2.6; I.2.7
dc.titleLexicalized Stochastic Modeling of Constraint-Based Grammars using Log-Linear Measures and EM Training
dc.typetext

Files

Collections

arXiv