Building a Hybrid Machine Translation System for Translating from English into Persian

Abstract

The incredible speed of contemporary modern world has made using machine translation/ MT an essential need. This paper aims at describing a hybrid machine translation system conducted for translation of some simple declarative English sentences into Persian. It is called a hybrid machine translation system because it employs a combination of both rule-based approaches and corpus-based approaches. The results of the research suggest that the given machine translation system produces more natural sentences which are compatible with Persian word order. Keywords: Machine translation system, Source language, Target language, Hybrid approach 1. Introduction Translation is one of necessities of the today's fast world. To meet this requirement in a faster manner, the human being thought of machine translation. The short history of machine translation has witnessed employing three main approaches in designing MT systems: rule-based approaches, corpus-based approaches, and hybrid approaches. Rule-based approaches and corpus-based approaches have been utilized in designing MT systems for translation from English into Persian more than hybrid approaches. Younesifar (1994) designed and built a machine translation system based on Augmented Transition Netwoks / ATN. The given system which used the sentence model of Wishon could produce a word for word translation of English sentences into Persian. Another machine translation system for translation from English into Persian is a rule-based machine translation based on HPSG. In this system, meaning is shown by Minimal Recursion Semantics / MRS semantic structure (Niknejad, 2008). Saedi (2008) introduced a hybrid machine translation for translation of simple English sentences into Persian. He (ibid) believed that disambiguation and transfer are the most important parts of this machine translation. Faroughi (2007) conducted a machine translation system based on lexical-functional grammar. The given system was designed for translating English sentences into Persian in general with a specific focus on translation of noun phrases. It seems that employing a hybrid approach would lead to generating better results. This research aims to show that utilizing a hybrid approach, i.e. using LFG as its rule-based approach and using a statistical parser as its corpus-based approach, in building an MT system for translating of some English one-place, two-place, and three-place predicators into Persian will produce acceptable and natural sentences. It is hoped that the findings of this research can be useful to translators, teachers, AI researchers and everyone else who needs an MT system.