132
click to vote
EMNLP
2007
15 years 4 months ago
2007
We use a generative history-based model to predict the most likely derivation of a dependency parse. Our probabilistic model is based on Incremental Sigmoid Belief Networks, a rec...