A Study on Divergence in Malayalam and Tamil Language in Machine Translation Perceptive

摘要

Machine Translation has made significant achievements for the past decades. However, in many languages, the complexity with its rich inflection and agglutination poses many challenges, that forced for manual translation to make the corpus available. The divergence in lexical, syntactic and semantic in any pair of languages makes machine translation more difficult. And many systems still depend on rules heavily, that deteriates system performance. In this paper, a study on divergence in Malayalam-Tamil languages is attempted at source language analysis to make translation process easy. In Malayalam-Tamil pair, the divergence is more reported in lexical and structural level, that is been resolved by using bilingual dictionary and transfer grammar. The accuracy is increased to 65 percentage, which is promising.