Dr. Amir Reza Shahbazkia, International Journal of Computer Science and Mobile Computing, Vol.8 Issue.4, April- 2019, pg. 270-311 © 2019, IJCSMC All Rights Reserved 270 Available Online at www.ijcsmc.com International Journal of Computer Science and Mobile Computing A Monthly Journal of Computer Science and Information Technology ISSN 2320–088X IMPACT FACTOR: 6.199 IJCSMC, Vol. 8, Issue. 4, April 2019, pg.270 – 311 Machine Translation by Homograph Detector with the Help of Grammatical Base of Persian Words 1 Dr. Amir Reza Shahbazkia 1 1105AmirReza@gmail.com Abstract: Language is core medium of communication and translation is core tool for the understand the information in unknown language. Machine translation helps the people to understand the information of unknown language without the help of Human translator. This study is brief introduction to machine Translation and the solution for homographs. machine translation have been developed for many popular languages and many researches and developments have been applied to those languages but a significant problem in Persian (the language of Iranian, Afghani, etc.) is detecting the homographs which is not generally problematic in any other languages except Arabic. Detection of homographs in Arabic have been extensively studied. However Persian and Arabic share 28 characters, having only 4 different characters, they are two quite different languages. Homographs, words with same spelling and different translations are more problematic to detect in Persian because not all the pronounced vowels are written in the text (only 20% of vowels are written in the text) so the number of homographs in Persian is about thousands of times more than in other languages except Arabic. In this paper we propose a new method for analysis and finding exact translation for homographs by algorithmic and grammatical rules. Keywords: homograph disambiguation, machine translation, Statistical, homograph disambiguation 1. Introduction significant problem in Persian (or Farsi) machine translation is homograph detection and disambiguation. This is not generally problematic in any other language except Arabic. Although a large work has been done for Arabic homograph detection and disambiguation with MADA [9], this work is useless for Persian. In fact Persian and Arabic are two quite different languages although they share 28 characters and have only 4 different ones. Since not all the vowels pronounced are actually written in the Persian and Arabic text, these two languages share a common problem in homograph detection and disambiguation but with different solutions. Moreover the number of homographs in Persian is about thousands of times more than in other languages, except Arabic. In Persian there are 32 characters from which 29 characters are consonants and the rest are vowels as shown below: بPronounced as b پPronounced as p ت طPronounced as t ث س صPronounced as s جPronounced as j چPronounced as ch=C ح هPronounced as h خPronounced as kh=x A