Automatic standardization of spelling variations of Hindi text

The phonetic nature of Indian languages and multiple dialects, transliteration of proper names, words borrowed from foreign languages has resulted in spelling variations of the same word. Such variations sometimes can be treated as errors in writing. While developing machine translation system, the task of standardizing the spellings for further processing the text is considered to play vital role in improving the accuracy of translation. In this paper, the rule based approach for standardizing spelling variations in the Hindi text while developing Hindi to Punjabi Machine Translation System has been explained. It was analyzed that only 7.45% text was standardized using this approach and thus had increased the accuracy of the machine translation system.