This paper describes an algorithm for exact decoding of phrase-based translation models, based on Lagrangian relaxation. The method recovers exact solutions, with certificates of optimality, on over 99% of test examples. The method is much more efficient than approaches based on linear programming (LP) or integer linear programming (ILP) solvers: these methods are not feasible for anything othe...

This paper describes an algorithm for exact decoding of phrase-based translation models, based on Lagrangian relaxation. The method recovers exact solutions, with certificates of optimality, on over 99% of test examples. The method is much more efficient than approaches based on linear programming (LP) or integer linear programming (ILP) solvers: these methods are not feasible for anything othe...

Using the modified factorization method employed in [5], we construct a new class of radial potentials whose spectrum for l = 0 coincides exactly with that of the hydrogen atom. A limiting case of our family coincides with the potentials previously derived by Abraham and Moses [6].

Training discriminative rule selection models is usually expensive because of the very large size of the hierarchical grammar. Previous approaches reduced the training costs either by (i) using models that are local to the source side of the rules or (ii) by heavily pruning out negative samples. Moreover, all previous evaluations were performed on small scale translation tasks, containing at mo...

The Study is the result of a collaboration between the SRM Agency and the University of Bologna – DEIS department, within the European Project EPTA. In this document are summarized analysis and results of the articles: Design and Control of a PublicTransportation Service Contract by A. Lodi, E. Malaguti, N. Stier-Moses e T. Bonino. Technical Report OR-12-2, DEIS – Università di Bologna and its ...

DFKI participated in the shared translation task of WMT 2015 with the GermanEnglish language pair in each translation direction. The submissions were generated using an experimental hybrid system based on three systems: a statistical Moses system, a commercial rule-based system, and a serial coupling of the two where the output of the rule-based system is further translated by Moses trained on ...