Hello all, Although not directly related to machine translation, you might be interested in a paper by R. C. Carrasco, M. A. Martínez-Prieto, J. Adiego and me on the compression of parallel texts that has been recently be published in the Journal of Artificial Intelligence Research: http://www.jair.org/papers/paper3500.html
The paper introduces the concept of generalized biwords to represent pairs of parallel words with high probability of co-occurrence that is able to describe reorderings and multi-word expressions. I think this may be inspiring to those working on the use of large parallel corpora in SMT. With kind regards -- Felipe Sánchez Martínez Dep. de Llenguatges i Sistemes Informàtics Universitat d'Alacant, E-03071 Alacant (Spain) Tel.: +34 965 903 400, ext: 2966 Fax: +34 965 909 326 http://www.dlsi.ua.es/~fsanchez _______________________________________________ Moses-support mailing list [email protected] http://mailman.mit.edu/mailman/listinfo/moses-support
