Strongly Incremental Constituency Parsing with
         Graph Neural Networks

          Kaiyu Yang                   Jia Deng
        Princeton University             Princeton University
      kaiyuy@cs.princeton.edu            jiadeng@cs.princeton.edu

                       Abstract
     Parsing sentences into syntax trees can benefit downstream applications in NLP.
     Transition-based parsers build trees by executing actions in a state transition system.
     They are computationally effi ...