Easy-First Dependency Parsing with Hierarchical Tree LSTMs
This provides a novel parsing method for NLP researchers and practitioners, though it appears incremental as it builds on existing tree LSTM and greedy parsing approaches.
The authors tackled dependency parsing by developing a compositional vector representation of parse trees using recursive combinations of recurrent-neural network encoders, achieving state-of-the-art accuracies for English and Chinese without external word embeddings.
We suggest a compositional vector representation of parse trees that relies on a recursive combination of recurrent-neural network encoders. To demonstrate its effectiveness, we use the representation as the backbone of a greedy, bottom-up dependency parser, achieving state-of-the-art accuracies for English and Chinese, without relying on external word embeddings. The parser's implementation is available for download at the first author's webpage.