DSpace DSpace UPC
  Pàgina principal | Llistar continguts | Cerca avançada | Com participar-hi Català   Castellano   English  


Títol: TAG, dynamic programming, and the perceptron for efficient, feature-rich parsing
Autor: Carreras Pérez, Xavier
Collins, Michael
Koo, Terry
Editorial: Coling 2008 Organizing Committee
Matèries: Àrees temàtiques de la UPC::Informàtica::Llenguatges de programació
Computational linguistics
Lingüística computacional
Tipus de document: Conference report
Descripció: We describe a parsing approach that makes use of the perceptron algorithm, in conjunction with dynamic programming methods, to recover full constituent-based parse trees. The formalism allows a rich set of parse-tree features, including PCFGbased features, bigram and trigram dependency features, and surface features. A severe challenge in applying such an approach to full syntactic parsing is the efficiency of the parsing algorithms involved. We show that efficient training is feasible, using a Tree Adjoining Grammar (TAG) based parsing formalism. A lower-order dependency parsing model is used to restrict the search space of the full model, thereby making it efficient. Experiments on the Penn WSJ treebank show that the model achieves state-of-the-art performance, for both constituent and dependency accuracy.
Altres identificadors i accés: Carreras, X.; Collins, M.; Koo, T. TAG, dynamic programming, and the perceptron for efficient, feature-rich parsing. A: Conference on Computational Natural Language Learning. "12th Conference on Computational Natural Language Learning (CoNLL-2011)". Manchester: Coling 2008 Organizing Committee, 2008, p. 9-16.
http://hdl.handle.net/2117/9806
Disponible al dipòsit:E-prints UPC
Comparteix:


SFX Query

Tots els ítems dipositats a UPCommons estan protegits per drets d'autor.

 

Valid XHTML 1.0! Programari DSpace Copyright © 2002-2004 MIT and Hewlett-Packard Comentaris
Universitat Politècnica de Catalunya. Servei de Biblioteques, Publicacions i Arxius