Phoneme Based Approach for Transliteration of Konkani Language
Sushma R. Iliger, Soniya Usgaonkar · International Journal of Computational and Electronic Aspects in Engineering · 2022
Simultaneously with the rise of machine translation, there has been a surge in the research field of machine transliteration. Despite the fact that the two processes are distinct and serve separate purposes, transliteration aids in the optimization of machine translation models. For languages such as Arabic, Korean, Japanese, Persian, Urdu, and Hindi etc, several methodologies have been developed. In this paper we present the implementation of the phoneme-based transliteration of Konkani scripture to Roman scripture using Direct Character Mapping technique and discuss the performance with respect to the ratings from a survey conducted from a small sample of Konkani speaking individuals. From the survey conducted we obtained an average score of 3.5 with respect to word accuracy.