Martínez Gómez, PascualSanchis Trilles, GermánCasacuberta Nolla, Francisco2014-02-142011978-3-642-21256-70302-9743https://riunet.upv.es/handle/10251/35666New variations on the application of the passive-aggressive algorithm to statistical machine translation are developed and compared to previously existing approaches. In online adaptation, the system needs to adapt to real-world changing scenarios, where training and tuning only take place when the system is set-up for the first time. Post-edit information, as described by a given quality measure, is used as valuable feedback within the passive-aggressive framework, adapting the statistical models on-line. First, by modifying the translation model parameters, and alternatively, by adapting the scaling factors present in stateof- the-art SMT systems. Experimental results show improvements in translation quality by allowing the system to learn on a sentence-by-sentence basis.8Reserva de todos los derechosOn-line learningPassive-aggressiveStatistical machine translationLENGUAJES Y SISTEMAS INFORMATICOSCapítulo de libro10.1007/978-3-642-21257-4_30Abierto