Talk from visiting scholar: Prof. Rafael Carrasco, University of Alicante

Prof. Rafael Carrasco, University of Alicante, Spain, http://www.dlsi.ua.es/~carrasco, is visiting CNGL/NCLT in DCU for a period of 4 weeks. As part of the CNGL/NCLT seminar series, he will give a talk tomorrow, Wednesday 01/08/2012, 16:00-17:00, L2.21, DCU School of Computing on:

– Statistical machine translation techniques for document translation retrieval and automatic modernisation of spelling.
We have compared different strategies to apply SMT techniques in order to retrieve documents which are a plausible translation of a given source document. In this apporach, both the probability of the translation and the relevance of the terms are taken into account in order to build an effective query. 

We also apply SMT to a Spanish diachronic corpus partially annotated with part-of-speech and modern form that we have released under open license. Experiments applying SMT to the modernisation of spelling with different training procedures report a character error rate below 0.30%, well below any of the baselines considered.

All are welcome to attend!