Identification and Translation of Idiomatic Sentence from Hindi to English

Abstract

This investigation is designed to shed light on identification and translation of idiomatic expression from Hindi to English language. The main problem encountered in idiomatic translation was investigated. Identification of idiom is utmost important resource for machine translation system and research. Machine translation system has been developed for many languages such as we have google translator, Bing translator etc. But they are failed to provide the correct translation of sentences containing idioms. The idiom parallel corpus was manually created to test the generated resource. The sentences containing idioms are translated with google translate system and noticed that it does not provide figurative meaning [1] which makes the translation of idioms rather difficult than any text translation. They are the real challenge for machine translation from the preliminary stage of machine translation development. A lot of research has been done for extraction and translation of text in many languages, but no significant research has been captured in Hindi to English idiom translation. In this paper we have given a rule based approach for the identification and translation of idiom using machine translation. The aim of a proper idiom translation is achieving equivalent sense and provide figurative meaning, strategies, cultural aspects and effects. The output is evaluated manually for intelligibility and accuracy. Further This Hindi to English idiom translation system can be expanded for other language pairs to improve their translation by encapsulating correct idiom translation with their ordinary translation.