A MULTI-ENGINE TRANSLATION APPROACH TO MACHINE TRANSLATION

One of the most difficult problems in dialogue machine translation is to correctly translate irregular expressions in natural conversations such as ungrammatical, incomplete, or ill-formed sentences. However, most existing machine translation systems reject utterances including irregular expressions. In this paper, we present a dialogue machine translation approach based on a cooperative distributed natural language processing model to attack the complex machine translation problem. In this approach, different types of translation processors are used in the analysis of the original language and the generation of the target language. The idea of combining multiple machine translation engines provides a new effective way to increase the success rate and quality of dialogue machine translation. A dialogue machine translation using multiple processors (DMTMP) system has been built using the following machine translation processors: (i) Robust Parser based Translation Processor, (ii) Example based Translation Processor, (iii) Family Modal based Translation Processor, and (iv) Super Function based Translation Processor. DMTMP is used in a practical machine translation environment called SWKJC. Experiments show that the approach presented in this paper is effective in implementation of robust dialogue machine translation systems.