Constituent parsing by classification

Abstract

Ordinary classification techniques can drive a conceptually simple constituent parser that achieves near state-of-the-art accuracy on standard test sets. Here we present such a parser, which avoids some of the limitations of other discriminative parsers. In particular, it does not place any restrictions upon which types of features are allowed. We also present several innovations for faster training of discriminative parsers: we show how training can be parallelized, how examples can be generated prior to training without a working parser, and how independently trained sub-classifiers that have never done any parsing can be effectively combined into a working parser. Finally, we propose a new figure-of-merit for best-first parsing with confidence-rated inferences. Our implementation is freely available at: http://cs.nyu.edu/~turian/software/parser/.