CIOREVIEW >> Artificial Intelligence >>

Machine Translations Transforming Modern Communication Patterns

By CIOReview | Friday, July 5, 2019

Travelling is fun but communicating isn't, and machine translation devices will transform the way we interact with people across the globe.

FREMONT, CA: Advanced technology in communication makes a significant impact on every industry. The potential information to be gained from watching and listening online is lost if the audience cannot understand the language of the lecturer. To solve this problem, scientists presented a solution with new machine learning. Machine translation systems have made it notably easy for someone to ask questions in a language never heard or seen before. The systems can also make innocent errors but achieve consistent communication for short exchanges, usually only a sentence or two long. Machine translation is an essential technology socio-politically, commercially and scientifically, despite multiple misconceptions about its progress or lack of it over the decades. The advent of the internet as one of the development media of modern communication has transformed translation into a bridge that unites speakers of various languages.

The endless traffic of interaction between distinctive language groups needs translation, but when on-the-spot translations are required, human translators aren't able to provide them fast enough. Translation using human translators is expensive and considerably slow when a large number of languages and subject fields are involved. To meet the growing translation requirement, machine translation systems are seen as a cost-effective option for human translators in a variety of situations. A research club obtained 46.5 hours of archived lecture visuals along with their transcriptions and English translations and demonstrated a deep learning-based practice to transcribe Japanese lecture speech to translate it into English. While seeing the videos, users would view subtitles in Japanese and English that matches the lecturer's speaking.

Measuring the quality of machine translation is a difficult challenge. Existing standardized quality scales provide a comparative and not absolute measure of quality. This is important because what's needed is an automated method to recognize problem texts so they can be routed for human analysis and post-edit. Human evaluation is pricey and time-consuming and thus irrelevant for frequent use during research and development of machine technology engines. Automatic translation quality evaluation plays an essential role in machine translation.