As evidenced by a number of machine translation competitions, statistical machine translation is producing encouraging results for language pairs where large corpora of previously translated texts are available for training. However, in practice the availability of such data is often a severe bottleneck. We therefore propose a methodology that only requires a bilingual dictionary and monolingual text corpora of the source and the target language, which should considerably relieve the data acquisition problem. What we suggest is a two stage procedure. In the first step we create a database of translation equivalents by extracting them from a pair of comparable monolingual corpora using a bilingual dictionary in combination with automatically generated thesauri of related words. In the second step we translate new sentences by retrieving appropriate translation equivalents from the database and by merging them using a combinatorial approach.
Field of science
- /humanities/languages and literature/general language studies
- /natural sciences/computer and information sciences/databases
Call for proposal
See other projects for this call