Abstract

This paper describes Picky, a probabilistic agenda-based chart parsing algorithm which uses a technique called probabilistic prediction to predict which grammar rules are likely to lead to an acceptable parse of the input. Using a suboptimal search method, Picky significantly reduces the number of edges produced by CKY-like chart parsing algorithms, while maintaining the robustness of pure bottom-up parsers and the accuracy of existing probabilistic parsers. Experiments using Picky demonstrate how probabilistic modelling can impact upon the efficiency, robustness and accuracy of a parser.

Citations

649	A stochastic parts program and noun phrase parser for unrestricted text - Church - 1988
594	An efficient context-free parsing algorithm - Earley - 1970
302	Self-Organized Language Modeling for Speech Recognition - Jelinek - 1990
148	Towards History-based Grammars: Using Richer Models for Probabilistic Parsing - Black, Jelinek, et al. - 1993
131	Recognition and parsing of context-free languages in time n 3 . Information and Control, 10(2):189–208. 42 Full Code for the Deductive Parsing Engine 〈Listing of file infer.pl〉 /*========================================================== Parser Based on a - Younger - 1967
83	An efficient recognition and syntax algorithm for contextfree algorithms - Kasami - 1965
54	Pearl: A Probabilistic Chart Parser - Magerman, Marcus - 1991
43	Poor estimates of context are worse than none - Gale, Church - 1990
40	Statistical Parsing of Messages - Chitrao, Grishman - 1992
35	Algorithm Schemata and Data - Kay - 1980
29	Generating a grammar for statistical training - Sharman, Jelinek, et al. - 1990
17	Efficient bottomup parsing - Moore, Dowding - 1991
15	Statistical agenda parsing - Bobrow - 1990
9	Probabilisti¢: Prediction and Picky Chart Parsing - Magerman, Weir - 1992