Xlit:A Tool for Transliteration between English and Indian Languages
Xlit is a transliteration tool to convert words from English to Indian languages and back, without losing the phonetic characteristics. It can be used in Machine Translation systems, e-governance applications and other applications that need to enter text in any Indian language and English. It can be used as an input method on Linux systems as well. Xlit adopts a statistical approach to tackle the transliteration problem.The transliteration problem is inherently ambiguous due to multiple variations of the same word, e.g. the word Raman may be transliterated as "रमन" and "रामन". Different spellings are used for same pronunciations e.g. शशि can be spelt as Shashi or as Sasi. Such variations are aplenty especially in the Indian context due to the rich linguistic diversity. The cases of acronyms and abbreviations pose another challenge as they need to be transliterated letter by letter e.g. CNBC needs to be transliterated as "सीएनबीसी". Given the inherent ambiguity of the problem, the tool generates multiple possible outputs for a given word.
Editor and OpenOffice.org Extension:
Xlit can be easily plugged into any application and invoked for performing transliteration. It can be used as an independent service by using our editor and also the downloadable OpenOffice.org Writer extension. The tool for English-Hindi language pair is available online (editor and extension). Other language pairs are available on request.
Please click here to download the Xlit OpenOffice.org Writer extension. Note that since both services, editor and extension, are under continued development, you may come across some instabilities / errors in the short run.
Please e-mail your valuable feedback and suggestions to us at kbcs@cdacmumbai.in. It will help us in improving the performance of the system.
Staff members associated with project Xlit: Dr. M Sasikumar, Chandrashekhar, Alok Dadhekar, Prakash Pimpale, Prashant More, Sarvesh Nikumbh, Aparna Mukherjee.
