Unsupervised spectral learning of finite-state transducers
Tipus de documentText en actes de congrés
Condicions d'accésAccés obert
Finite-State Transducers (FST) are a standard tool for modeling paired inputoutput sequences and are used in numerous applications, ranging from computational biology to natural language processing. Recently Balle et al. presented a spectral algorithm for learning FST from samples of aligned input-output sequences. In this paper we address the more realistic, yet challenging setting where the alignments are unknown to the learning algorithm. We frame FST learning as finding a low rank Hankel matrix satisfying constraints derived from observable statistics. Under this formulation, we provide identifiability results for FST distributions. Then, following previous work on rank minimization, we propose a regularized convex relaxation of this objective which is based on minimizing a nuclear norm penalty subject to linear constraints and can be solved efficiently.
CitacióBailly, R.; Carreras, X.; Quattoni, A.J. Unsupervised spectral learning of finite-state transducers. A: Neural Information Processing Systems Conference. "Advances in Neural Information Processing Systems 26". Lake Tahoe, Nevada: 2012, p. 800-808.