Transition-based Natural Language Processing
Transition-based models use a transition system, or abstract state machine, to model structured prediction problems, for example syntactic dependency parsing, as a sequence of actions. We propose a technique for learning representations of these states.…