Graphical model showing the factorization of our noisy channel model where yi indicates the ith target language sentence and xi indicates the ith source language sentence. In the prior (top) the target sentences (the yi’s) only influence the corresponding source sentence and therefore can be learned and modeled independently, but at test time (bottom), when the target is not observed, each yi depends on every xi.
This site uses cookies. By continuing to use our website, you are agreeing to our privacy policy. No content on this site may be used to train artificial intelligence systems without permission in writing from the MIT Press.