Paper: Dependency Parsing with Energy-based Reinforcement Learning

ACL ID W09-3838
Title Dependency Parsing with Energy-based Reinforcement Learning
Venue International Conference on Parsing Technologies
Session Main Conference
Year 2009
Authors

We present a model which integrates dependency parsing with reinforcement learning based on Markov decision pro- cess. At each time step, a transition is picked up to construct the dependency tree in terms of the long-run reward. The op- timal policy for choosing transitions can be found with the SARSA algorithm. In SARSA, an approximation of the state- action function can be obtained by calcu- lating the negative free energies for the Restricted Boltzmann Machine. The ex- perimental results on CoNLL-X multilin- gual data show that the proposed model achieves comparable results with the cur- rent state-of-the-art methods.