A NOVEL APPROACH TO MACHINE TRANSLATION: A PROPOSED LANGUAGE-INDEPENDENT SYSTEM BASED ON DEDUCTIVE SCHEMES

Abstract

Compared to corpora-based machine translation methods, rule-based methods havedeficiencies, which make them unattractive for the researchers of this field. The first problem isthat these methods are language dependent. Rule-based methods require the syntactic informationabout source and target languages. On the other hand, in many cases, especially for proverbs andspecific expressions, syntactic rules are not useful anymore. In such cases, the use of examplebasedapproaches is inevitable. In this work, we propose and integrate a set of novel schemes tointroduce a new translation system, called BORNA. First a grammar induction method based onthe Expectation Maximization (EM) algorithm is proposed. After representing the extractedknowledge in the form of a set of nested finite automata, a recursive model is proposed, whichuses a combination of rule and example based techniques. In the translation phase, through ahierarchical chunking process, the input sentence is divided into a set of phrases. Each phrase issearched in the corpus of examples. If the phrase is found, it will not be chunked anymore.Otherwise, the phrase is divided into smaller sub-phrases. The simulation results show thatBORNA outperforms its counterparts, significantly. Compared to PARS, Frengly and Googletranslators, BORNA receives the highest Bleu scores for its translations, while it results in theminimum values for different error measures, including PER, TER and WER.