Incremental Parsing with Minimal Features Using Bi-Directional LSTM

• DOI: 10.18653/v1/p16-2006
• Publication Date: 2016-08-13T19:45:39Z
• Authors: James Cross; Liang Huang
• Keywords: FOS: Computer and information sciences; Computer Science - Computation and Language; 0202 electrical engineering, electronic engineering, information engineering; 02 engineering and technology; Computation and Language (cs.CL)

Recently, neural network approaches for parsing have largely automated the combination of individual features, but still rely on (often a larger number of) atomic features created from human linguistic intuition, and potentially omitting important global context.To further reduce feature engineering to bare minimum, we use bi-directional LSTM sentence representations model parser state with only three positions, which automatically identifies aspects entire sentence.This achieves state-of-the-art results among greedy dependency parsers English.We also introduce novel transition system constituency does not require binarization, together above architecture, both English Chinese.