An Analysis of Phrase based SMT for English to Manipuri Language

Abstract

Statistical Machine Translation (SMT) is one ruling approach adopted for developing major translation systems today. Here, we report a phrase-based SMT system from English to Manipuri. The variance in the structure and morphology between English and Manipuri languages and the lack of resources for Manipuri languages pose a significant challenge in developing an MT system for the language pair. In comparison, English has poor morphology and SVO structure and belongs to the Indo-European family. Manipuri language has richer morphology and SOV structure and belongs to the Sino-Tibetan family. Manipuri has two scripts- Bengali script and Meitei script. Here the Bengali script is used for developing the system. Our system uses the Moses toolkit. We train and test the system using the tourism, agriculture and entertainment corpus. Further, we use the BLEU metric to evaluate the systems' performance.