Daniel Quernheim

Learn More
We present a new translation model integrating the shallow local multi bottom-up tree transducer. We perform a large-scale empirical evaluation of our obtained system, which demonstrates that we significantly beat a realistic tree-to-tree base-line on the WMT 2009 English → German translation task. As an additional contribution we make the developed(More)
Minimal deterministic finite automata (dfas) can be reduced further at the expense of a finite number of errors. Recently, such minimization algorithms have been improved to run in time O(n log n), where n is the number of states of the input dfa, by [Gawrychowski and Jeż: Hyper-minimisation made efficient. Proc. Mfcs, Lncs 5734, 2009] and [Holzer and(More)
Weighted finite-state acceptors and transducers (Pereira and Riley, 1997) are a critical technology for NLP and speech systems. They flexibly capture many kinds of stateful left-to-right substitution, simple transducers can be composed into more complex ones, and they are EM-trainable. They are unable to handle long-range syntactic movement, but tree(More)
Hyper-minimisation of deterministic nite automata is a recently introduced state reduction technique that allows a nite change in the recognised language. A generalisation of this lossy compression method to the weighted setting over semields is presented, which allows the recog-nised formal power series to dier for nitely many input strings. First, the(More)
An open question in [F ¨ UL¨OP, MALETTI, VOGLER: Weighted extended tree transducers. Fundamenta Informaticae 111(2), 2011] asks whether weighted linear extended tree transducers preserve recogniz-ability in countably complete commuta-tive semirings. In this contribution, the question is answered positively, which is achieved with a construction that(More)
We present an experimental statistical tree-to-tree machine translation system based on the multi-bottom up tree transducer including rule extraction, tuning and decoding. Thanks to input parse forests and a " no pruning " strategy during decoding , the obtained translations are competitive. The drawbacks are a restricted coverage of 70% on test data, in(More)
A composition procedure for linear and nondeleting extended top-down tree transducers is presented. It is demonstrated that the new procedure is more widely applicable than the existing methods. In general, the result of the composition is an extended top-down tree transducer that is no longer linear or nondeleting, but in a number of cases these properties(More)