ANN and Rule Based Model for English to Sanskrit Machine Translation
Vimal Kant Mishra, R. B. Mishra · Americanae (AECID Library) · 2009
Abstract. The development of Machine Translation system for ancient language such as Sanskrit language is much more fascinating and challenging task. Due to lack of linguistic community, there are no wide work accomplish in Sanskrit translation while it is mother language by virtue of its importance in cultural heritage of India. In this paper, we integrate a traditional rule based approach of machine translation with Artificial Neural Network (ANN) model which translates an English sentence (source language sentence) into equivalent Sanskrit sentence (target language sentence). We use feed forward ANN for the selection of Sanskrit word like noun, verb, object, adjective etc from English to Sanskrit User Data Vector (UDV). Due to morphological richness of Sanskrit language, this system makes limited use of syntax and uses only morphological markings to identify Subject, Object, Verb, Preposition, Adjective, Adverb and as well as Conjunctive sentences also. It uses limited parsing for part of speech (POS) tagging, identification of clause, its Subject, Object, Verb etc and Gender-Number-Person (GNP) of noun, adjective and object. This system represents the translation between the SVO and SOV classes of languages. This system gives translation result in GUI form and handles English sentences of different classes.