Query Word Labeling using Supervised Machine Learning

We have participated in the first sub task of FIRE 2014 Shared Task on Transliterated Search where the Indian language is Hindi. The aim of this task is to identify words as belonging to an Indian language (L) or English (E) from sentences written in Roman script and if the word belongs to Indian language (L), transliterate the same to its Devanagari script equivalent. We have used a supervised machine learning approach to classify words as belonging to either L or E and a rule-based approach for transliterating words by first splitting them into consonant or vowel blocks.