GRT: Gurmukhi to Roman Transliteration System using Character Mapping and Handcrafted Rules

Abstract

In the last two decades, the transliteration system has got significant research attention. It is observed that Punjabi to English transliteration for all type of part-of-speech words is comparably less studied. Currently, some research work in this area is carried out but only for proper nouns and some technical terms. So, there is need to focus on all type of words. The Gurmukhi to Roman transliteration (GRT) system is the first proposed system for transliteration of all kind of part-of-speech words. This system uses the handcrafted-rules and character mapping (CM) approach for transliteration between languages involved. The CM is done for Gurmukhi script with its equivalent to Roman script. It transliterates text written in Punjabi language into English language. It is tested on 65,130 Punjabi words and achieved accuracy of 99.27%, which is better than other state-of-art system results. The developed system can be used in social media text normalization, translation, sentiment analysis of multilingual text, text summarization of multilingual text, etc.