Third-order Variational Reranking on Packed-Shared Dependency Forests

We propose a novel forest reranking algorithm for discriminative dependency parsing based on a variant of Eisner’s generative model. In our framework, we define two kinds of generative model for reranking. One is learned from training data offline and the other from a forest generated by a baseline parser on the fly. The final prediction in the reranking stage is performed using linear interpolation of these models and discriminative model. In order to efficiently train the model from and decode on a hypergraph data structure representing a forest, we apply extended inside/outside and Viterbi algorithms. Experimental results show that our proposed forest reranking algorithm achieves significant improvement when compared with conventional approaches.