Human vs Machine Translation Infographic: And the Winner is...

Human vs Machine Translation Infographic: And the Winner is...

How does a human translator compare against modern machine translators?

See the infographic below this article.

How far can we go to replacing human translators with artificial ones?

The Study

As translators ourselves, here at VerbalInk we were extremely interested in Google’s latest translation efforts. Would the machine live up to the human version? Google’s new software app promised visual and audio translation that could be used in conjunction with video conferencing services, similar to the very popular Skype.

So we enlisted some of our best translators to test the machine. Could the machine grasp sarcasm, colloquial jargon and other tiny nuances that make translation difficult even for the most skilled translator?

Our two Spanish translators, Adriana and Gaby, were enlisted to take on the machine. Two tests were prepared to determine a winner. The first test required Adriana to translate a marketing document from the National Beekeeping Council. The second test to convert a minute of audio from Spanish to English. Gaby, our quality control specialist, acted as the judge. After reviewing the document and the audio recording, Gaby gave us the results on three categories: word choice, grammar and syntax and accuracy and comprehension.

The Results

While Google Translate held up valiantly against Adriana, it simply could not perform at the same level as our professional translators. For basic translation and comprehension, Google passed the test. Gaby could understand Google’s interpretation for both the written and audio translations; however, had Google been a real person, she would have been considered a novice English speaker, at best.

Google’s translations produced raw material that lacked proper grammar and nuanced word choices. Google’s interpretations of the text and the audio sounded like they had been input into a computer and spit out at the reviewer.

Adriana’s translation, on the other hand, was far more superior to Google’s. She was able to convey the basic messages of both the audio and visual documents in a clear, concise way. She was able to convey tone, style and subject matter. Plus, she sounded like a real human being — something computers have yet to accomplish.

Conclusion

While translation programs are attractive to consumers and much lower in cost than human translators, they simply don’t live up to the real-live version. Though Google is headed in the right direction and may be able to replicate a human in the not-so distant future, their current attempts sadly don’t live up to their smart cars and glasses. In this day and age, a computer simply cannot live up to a human’s empathy and understanding of syntax, style and culture.

The Infographic

Share

Free Language Experience

Learn to create experiences that teach you a new language

Related Resources