Decision tree models applied to the labeling of text with parts-of-speech
Ezra W. Black, Frederick Jelinek, John Lafferty, Robert L. Mercer, Salim Roukos · 1992
We describe work which uses decision trees to estimate marginal probabilities in a maximum entropy model for predicting the part-of-speech of a word given the context in which it appears. Two experiments are presented which exhibit improvements over the usual hidden Markov model approach.