A phoneme based model for english to malayalam transliteration

Abstract

Transliteration plays an important role nowadays because it can be used for many applications such as text to speech conversion, Dubbing etc. The main purpose of Transliteration is to convert words from one language to another language such that their pronunciation will be same, but the meaning may be different. Transliteration is mainly classified into 4 models such as Phoneme based Transliteration Model, Correspondence based Transliteration Model, Grapheme based Transliteration Model, and Hybrid based Transliteration Model. Phoneme based transliteration takes the advantage of pronunciation and is better to implement when compared to other models. Research works are taking place in this field especially with Korean, Arabic, Chinese & Indian Languages. Very little work is being carried out with English to Malayalam Transliteration. At present Grapheme based transliteration is available for English - Malayalam transliteration. But there is a problem of pronunciation. This paper presents a Phoneme based transliteration model for English - Malayalam conversion which is based on pronunciation and works better than the Grapheme based model for the same. This model uses a pronunciation dictionary.