Results 1 
2 of
2
Statistical inference and probabilistic modeling for constraintbased NLP
 Computers, Linguistics, and Phonetics between Language and Speech: Proceedings of the 4th Conference on Natural Language Processing (KONVENS’98
, 1998
"... In this paper we present a probabilistic model for constraintbased grammars and a method for estimating the parameters of such models from incomplete, i.e., unparsed data. Whereas methods exist to estimate the parameters of probabilistic contextfree grammars from incomplete data ([2]), so far for ..."
Abstract

Cited by 6 (2 self)
 Add to MetaCart
(Show Context)
In this paper we present a probabilistic model for constraintbased grammars and a method for estimating the parameters of such models from incomplete, i.e., unparsed data. Whereas methods exist to estimate the parameters of probabilistic contextfree grammars from incomplete data ([2]), so far for probabilistic grammars involving contextdependencies only parameter estimation techniques from complete, i.e., fully parsed data have been presented ([1]). However, completedata estimation requires laborintensive, errorprone, and grammarspecific handannotating of large language corpora. We present a loglinear probability model for constraint logic programming, and a general algorithm to estimate the parameters of such models from incomplete data by extending the estimation algorithm of [8] to incomplete data settings. Diese Arbeit präsentiert ein probabilistisches Modell für kontextsensitive, constraintbasierte Grammatiken und erstmals eine Methode, die Parameter solcher probabilistischer Modelle anhand unvollständiger Daten einzuschätzen. Probabilistische Grammatiken werden hier in einem loglinearen
Probabilistic Constraint Logic Programming
, 1997
"... . This paper addresses two central problems for probabilistic processing models: parameter estimation from incomplete data and efficient retrieval of most probable analyses. These questions have been answered satisfactorily only for probabilistic regular and contextfree models. We address these pro ..."
Abstract
 Add to MetaCart
. This paper addresses two central problems for probabilistic processing models: parameter estimation from incomplete data and efficient retrieval of most probable analyses. These questions have been answered satisfactorily only for probabilistic regular and contextfree models. We address these problems for a more expressive probabilistic constraint logic programming model. We present a loglinear probability model for probabilistic constraint logic programming. On top of this model we define an algorithm to estimate the parameters and to select the properties of loglinear models from incomplete data. This algorithm is an extension of the improved iterative scaling algorithm of Della Pietra, Della Pietra, and Lafferty (1995). Our algorithm applies to loglinear models in general and is accompanied with suitable approximation methods when applied to large data spaces. Furthermore, we present an approach for searching for most probable analyses of the probabilistic constraint logic prog...