Stochastic Parse-Tree Recognition by a Pushdown Automaton
Frédéric
Tendeau
author
1995-sep 20-24
text
Proceedings of the Fourth International Workshop on Parsing Technologies
Eva
Hajicova
editor
Bernard
Lang
editor
Robert
Berwick
editor
Harry
Bunt
editor
Bob
Carpenter
editor
Ken
Church
editor
Aravind
Joshi
editor
Ronald
Kaplan
editor
Martin
Kay
editor
Makoto
Nagao
editor
Anton
Nijholt
editor
Mark
Steedman
editor
Henry
Thompson
editor
Masaru
Tomita
editor
K
Vijay-Shanker
editor
Yorick
Wilks
editor
Kent
Wittenburg
editor
Association for Computational Linguistics
Prague and Karlovy Vary, Czech Republic
conference publication
We present the stochastic generalization of what is usually called correctness theorems: we guarantee that the probabilities computed operationally by the parsing algorithms are the same as those defined denotationally on the trees and forests defined by the grammar. The main idea of the paper is to precisely relate the parsing strategy with a parse-tree exploration strategy: a computational path of a parsing. algorithm simply performs an exploration of a parse-tree for the input portion already parsed. This approach is applied in particular to Earley and Left-Corner parsing algorithms. Probability computations follow parsing operations: looping problems (in rule prediction and subtree recognition) are solved by introducing probability variables (which may not be immediately evaluated). Convergence is ensured by the syntactic construction that leads to stochastic equations systems, which are solved as soon as possible. Our algorithms accept any (probabilistic) CF grammar. No restrictions are made such as prescribing normal form, proscribing empty rules or cyclic grammars.
tendeau-1995-stochastic
https://aclanthology.org/1995.iwpt-1.28
1995-sep 20-24
234
249