Conditional Structure versus Conditional Estimation in NLP Models

Abstract

This paper separates conditional parameter estima- tion, which consistently raises test set accuracy on statistical NLP tasks, from conditional model struc- tures, such as the conditional Markov model used for maximum-entropy tagging, which tend to lower accuracy. Error analysis on part-of-speech tagging shows that the actual tagging errors made by the conditionally structured model derive not only from label bias, but also from other ways in which the independence assumptions of the conditional model structure are unsuited to linguistic sequences. The paper presents new word-sense disambiguation and POS tagging experiments, and integrates apparently conflicting reports from other recent work.

Links

PhilArchive



    Upload a copy of this work     Papers currently archived: 74,310

External links

Setup an account with your affiliations in order to access resources via your University's proxy server

Through your library

  • Only published works are available at libraries.

Similar books and articles

Conditional Intentions.Luca Ferrero - 2009 - Noûs 43 (4):700 - 741.
Revision Algebra Semantics for Conditional Logic.John Pais - 1992 - Studia Logica 51 (2):279 - 316.
Conditional Desirability.Richard Bradley - 1999 - Theory and Decision 47 (1):23-55.
Iterated Revision and Minimal Change of Conditional Beliefs.Craig Boutilier - 1996 - Journal of Philosophical Logic 25 (3):263 - 305.

Analytics

Added to PP
2010-12-22

Downloads
10 (#881,601)

6 months
1 (#415,900)

Historical graph of downloads
How can I increase my downloads?

Author's Profile

Daniel Klein
Harvard University

References found in this work

No references found.

Add more references