UNL Enconversion framework for Machine Translation

This paper presents the building of a language independent Enconversion framework of UNL. Universal Networking Language is a relatively new interlingua of machine translation which is an important application of NLP .The purpose of framework is to localize the software applications and platforms and thus trying to bridge the great digital divide that has created due the linguistic barriers.It is designed to perform the task of converting source languages into UNL format; i.e. UNL expressions. The process of building the framework of Enconverter system includes the task of identifying the category of words in a corpus by doing the morphological analysis. These tagged words are used as input by the UNL Enconverter system to find the semantic relation existing between the words and to generate UNL relations . Enconverter framework maps all the 46 UNL relations and then tested with Dravidian language Malayalam . Later the Enconversion module is altered to include the global language English also.A resultant hyper graph is also generated with concepts as the nodes and arch connecting them showing the relation existing between them.The language corpora once converted to UNL, can be translated to any other language given UNL system built for that language.