Abstract

This paper points out that the Penn lI treebank representations are of the kind predicted to have such an effect, and describes a simple node relabeling transformation that improves a treebank PCFG-based parser's average precision and recall by around 8%, or approximately half of the performance difference between a simple PCFG model and the best broad-coverage parsers available today. This performance variation comes about because any PCFG, and hence the corpus of trees from which the PCFG is induced, embodies independence assumptions about the distribution of words and phrases. The particular independence assumptions implicit in a tree representation can be studied theoretically and investigated empirically by means of a tree transformation / detransformation process

Citations

1654	B.: Building a large annotated corpus of english: The Penn treebank - Marcus, Marcinkiewicz, et al. - 1993
537	Neural networks and the bias/variance dilemma - Geman, Bienenstock, et al. - 1992
496	Statistical Language Learning - Charniak - 1993
396	New Statistical Parser Based on Bigram Lexical De-pendencies - Collins - 2006
324	Statistical Parsing with a Context-Free Grammar and Word Statistics - Charniak - 1997
315	Empirical Methods for Artificial Intelligence - Cohen - 1995
265	Structural ambiguity and lexical relations - Hindle, Rooth - 1993
203	Tree-Bank Grammars - Charniak - 1996
94	Bracketing Guidelines for Treebank II Style Penn Treebank Project - Bies - 1995
75	Neural networks and the bias/variance dilemma. Neural Comput - Geman, Bienenstock, et al. - 1992
54	Pearl: A Probabilistic Chart Parser - Magerman, Marcus - 1991
43	Estimation of probabilistic context-free grammars - Chi, Geman - 1998
40	Context-sensitive statistics for improved grammatical language models - Systems, Charniak, et al. - 1994
37	The Theory of Parsing, Translation and Compiling, volume 1: Parsing - Aho, Ullman - 1972
29	Compacting the Penn Treebank grammar - Krotov, Gaizauskas, et al. - 1998
26	The complexity of the vocabulary of Bambara - Culy - 1985
12	The Complexity of the Vocabulary - Culy - 1985
9	Bracketting Guideliness for Treebank II style Penn Treebank Project. Linguistic Data Consortium - Bies, Ferguson, et al. - 1995