Bat-Erdene Batsukh

¹School of Information and Communication Management, University of the Humanities, Ulaanbaatar, Mongolia

Correspondence: Bat-Erdene Batsukh, School of Information and Communication Management, University of the Humanities, Sukhbaatar square 20, Ulaanbaatar, Mongolia.

Tel: 976-8805-9362. E-mail: [email protected]

ABSTRACT

In this globalized world, translation plays a major role in bridging the gap between people. The need for translation arises when a person, regardless of where they are in the world, needs to understand, study, and express content in a language other than their native language. With the increasing number of social media platforms, users are increasingly exposed to content written in foreign languages ​​by other users, and researchers are increasingly required to conduct research and analysis in multiple languages ​​and publish their results in foreign languages, the need for real-time translation.

Figure 1. “UH TRANSLATE” real time multilingual neural machine translator

Professional translation is labor-intensive and expensive, so automatic translation, also known as machine translation, plays a crucial role in helping millions of users understand content in foreign languages. Machine translation can be used not only for individual translations by ordinary users, but also to help professional translators translate faster. We have successfully developed and launched version 1.0 of our multilingual translation program called “UH TRANSLATE” (Figure 1). Until now, Mongolian-English and English-Mongolian translations have been completely dependent on Google Translate and other internet sources. The feature of this program is that the translation engine (UH TRANSLATE) and the application interface (UTRANSLATE) all work independently.

Video 1. “UH TRANSLATE” translation workflow

This translation program that we have developed has the following advantages (Video 1).

o Mongolian-English, English-Mongolian, English-Greek, English-German translation

o Multilingual translation

o Real-time translation

o File translation

o Future addition and expansion of any language translation

We have trained Mongolian-English, English-German and English-Greek translation models with the data prepared within the framework of the ARTEST project in collaboration with the exchange students (University of the Humanities, Otgontenger University, University of Cologne, University of the Peloponnese and International Hellenic University).

This article was written by Bat-Erdene Batsukh as a result article for ARTEST student exchange activities.