Hybrid Machine Translation System for the Translation of Simple English Prepositions and Periphrastic Causative Constructions from English to Hindi

Abstract

Machine Translation (MT) is one of the applications of Computational Linguistics (CL), Artificial Intelligence (AI) and Natural Language Processing (NLP).When a text in one human language is translated into another by a machine, it is called Machine Translation (MT). All natural languages are highly complex, and each has its own language specificity. Most of the lexical items in all natural languages exhibit polysemy and synonymy. The syntactic complexity of languages, the variations in syntactic constructions found in the languages, different semantics of same and variant syntactic constructions and the polysemic and synonymic nature of lexical items of natural languages still remain unresolved problems in the context of Machine Translation (MT). Translation is a series of steps, which involves the replacement of human thoughts expressed in one natural language elements (Source Language/SL) into its equivalent natural language elements in another natural language (Target Language/TL). The process of translation involves: Comprehension: understanding and interpretation of spoken and written language Formulation: putting together the linguistic elements in appropriate syntax to express the context of the source text. The polysemic and synonymic nature of the lexical items of the source and the target languages make them highly expressive. The lexical items are classified as functional words and content words. The major part of the lexicon contains the content words. The functional words are few in number or they are countable. The functional words play a vital role in determining the syntax and semantics of the sentences in a particular natural language. The functional words like pre/post positions, auxiliary verbs and helping verbs are highly polysemic and synonymic in nature. They also perform many grammatical functions and build different types of sentences. They show one- to-many and many-to-one translation equivalence in different languages..