Translating Persian documents into English using knowledge based WSD

The necessity of machine translation systems is growing rapidly due to the increase of documents and translation requests. Reading web pages, news, articles, manuals and users' guides and getting the gist of a test are some of the examples for our daily need to translation. This paper introduces a Persian to English machine translation system, named PEnT2. It uses a transfer based approach and employs the grammatical role of the sentence words as the main clue to perform the translation processes. PEnT2 translates simple Persian sentences into English using a hybrid approach. It exploits the advantages of rule based, knowledge based and corpus based methods in different components of a machine translation system including word sense disambiguation, structural transfer and structure optimization. Experiments show improved results in comparison with other available systems.