A SURVEY ABOUT MACHINE TRANSLATION OF TURKIC LANGUAGES

With developing Internet technology, it becomes important tasks to write and edit documents through the Internet supported tools. In such tasks, spell checking, auto completing words after some syllables, machine translation of words from one language to another, content summarizing and analyzing according to words and sentence structure in a file, are considered popular and important topics. With improving social communication technology and devices, analyzing continuously produced documents and translation between different languages are becoming more important day after day. It is difficult to handle such task with manually without any error and with high speed. Because of these reasons, computer supported research languages come into prominence and a new filed, which is computational linguistic, formed in science. Turkish is one of the most computational researched language in Turkic language family and a lot of projects have been done. For other Turkic languages, computational researches are still at the beginning stage. This paper describes some main properties of the Turkic languages and summarize machine translations and natural language researches that have been done so far. A unique tagging system for Turkic word is suggested to implement a machine translations system between all Turkic languages, even between other Altaic languages. The usability of the unique tags is explained with examples.