Spectral Learning for Non-Deterministic Dependency Parsing

Franco M. Luque


20 Abril 2012

11:00h - Arribada dels assistents i estona de germanor

11:15h - Presentació


Joint work with Ariadna Quattoni, Borja Balle and Xavier Carreras

In this talk we will present our EACL-2012 paper. We study spectral learning methods for non-deterministic split head-automata grammars, a powerful hidden-state formalism for dependency parsing. We present a learning algorithm that, like other spectral methods, is efficient and non-susceptible to local minima. We show how this algorithm can be formulated as a technique for inducing hidden structure from distributions computed by forward-backward recursions. Furthermore, we also present an inside-outside algorithm for the parsing model that runs in cubic time, hence maintaining the standard parsing costs for context-free grammars.

A copy of the EACL paper can be obtained from: http://www.lsi.upc.edu/~carreras/pubs/2012-eacl-lqbc.pdf

Scroll to Top