Start page
Call for papers [pdf]
Paper submission
Instructions for the presenters

Registration closed


SLTC 2008

November 20 - 21, 2008

Statistical machine translation: Where do we stand?

Hermann Ney

RWTH Aachen University, Aachen, Germany

When statistical machine translation (MT) was introduced 20 years ago, it was considered by many experts to be a singularity. Today it is the mainstream direction for research on MT. The last 20 years have shown a steady progress in statistical MT which parallels many developments in statistical speech recognition. Today's statistical MT systems are able to translate across a wide range of language pairs. In this talk, we review the basic components of statistical MT: training of alignment and lexicon models, extraction of bilingual phrases of word groups, log-linear modelling for hypothesis scoring and search, i.e. generation of the target sentence. We will discuss the progress that the field has made so far and address the present shortcomings and the future challenges.

Slides from the presentation

Important dates

June 1, 2008:
Deadline workshop proposals
September 15, 2008
Deadline papers
October 30, 2008
Deadline of revised abstract
November 6, 2008
Deadline registration
November, 20-21