Evaluation of dynamic bayesian network models for entity name transliteration

Abstract

This paper proposes an evaluation of DBN models so as to identify DBN configurations that can improve machine transliteration accuracy