A New Model for Connectionist Machine Translation
Xu Luomai, Ph.D.
Guangdong University of Foreign Studies
This thesis aims at developing a connectionist technique for machine translation (MT).
This technique treats MT development as seeking the solutions to two sub-problems. The first
problem concerns obtaining vocabulary translations and is solved by using a distributed
neural translation lexicon. The second problem involves adjusting the transliterations
produced by the lexicon into acceptable target language sentences, which is handled by a
The translation lexicon learns the meaning of words from exampIes and stores the
acquired lexical knowledge in a set of lexical networks. During translation, the lexical
networks perform lexical disambiguation automatically. With the neural lexicon,
programming a disambiguation component for an MT system is no longer necessary. The
technique allows a translation lexicon to scale up easily to a full size lexicon for an MT
application. Although developed initially for English-Chinese translation, the technique can
be used to develop translation lexicons between any language pairs.
The hybrid generator consists of a generation network (GN) and a symbolic generator
(SG). GN learns a simple jumble of grammar from examples, and SG physically adjusts
transliterations to produce turget language sentences. The generator is a bold attempt at
language generation without using any formal linguistic theories. This thesis discusses its
strengths and weaknesses. This new language generation technique is still under development,
requiring further research before becoming a practical alternative to those based on the