Abstract

We study the application of Hidden Markov Models (HMMs) to learning information extractors for ¤-ary relations from free text. We propose an approach to representing the grammatical structure of sentences in the states of the model. We also investigate using an objective function during HMM training which maximizes the ability of the learned models to identify the phrases of interest. We evaluate our methods by deriving extractors for two binary relations in biomedical domains. Our experiments indicate that our approach learns more accurate models than several baseline approaches. 1

Citations

3116	A tutorial on hidden Markov models and selected applications in speech recognition - Rabiner - 1989
1459	An algorithm for suffix stripping - Porter - 1980
422	An Inequality and Associated Maximization Technique in Statistical Estimation of a Markov Process - Baum - 1972
355	Maximum entropy Markov models for information extraction and segmentation - McCallum, Freitag, et al. - 2000
147	Estimating Probabilities: A Crucial Task in Machine Learning - Cestnik - 1990
128	Rosenreid R. Learning hidden Markov model structure for information extraction - Seymore, McCallum - 1999
111	Information Extraction with HMMs and Shrinkage - Freigtag, McCallum - 1999
89	Information Extraction with HMM Structures Learned by Stochastic Optimization - Freitag, McCallum - 2000
76	Information extraction using hidden Markov models. Master’s thesis - Leek
73	An empirical study of automated dictionary construction for information extraction in three domains - Riloff - 1996
9	Yeast protein database (YPD): a database for the complete proteome of Saccharomyces cerevisiae - Payne, Garrels - 1997
9	markov models for labeled sequences - Hidden - 1994
3	The sundance sentence analyzer - Riloff - 1998