Rule based hardware approach for machine transliteration: A first thought

摘要

It is very challenging to translate proper noun across intra linguistic alphabets. It may not transliterate properly due to insufficient knowledge of the system. Moreover if we translate with supervise approach then overall speed will deteriorate at the time of training and testing due to large search space and searching time which is very challenging to reduce. In this manuscript, we propose an alternative approach to reduce the overall time of conversion compare to existing software approach. This can be achieve with hardware implementation. The hardware implementation for transliteration is the first time in the literature.