How AI/Machine learning works in translator

Question

I am an undergraduate community college student in Richmond, VA.

I am wondering if you can help me understand how the translator works or if you can connect me with another expert who might be able to help me learn exactly how google translate works.

Questions:

If you enter the same phrase into the translator multiple times, does the program learn from it? If so, how?

Does the translator take what people enter into the translator, store it and use it to output a common translation?

Here is a scenario. Let's say I wrote something in english and then translated it word for word into farsi. Then, I took my farsi work, inputed into google translate, and the output was a translation almost word for work with my english translation. IF I keep doing that, would the translator "deep learn" and correct itself so that it exactly matched my english translation? What about if I kept hitting the reverse arrow, and changing a word here or there, until both the farsi and english sentences inside the translator, match exactly what I originally wrote? Would the translator learn from that? In that way could I have "taught" the translator the most common way to interpret sentences in either language? Is there a way to prove that someone used a translator to write something? IF someone is trying to prove that someone else used a translator to create a literary work, instead of writing it themselves, can they fairly say that the translator's translation matching word for word is indisputable, undeniable proof of their accusation?

Very Respectfully,

Answer 1

Hi, thanks for reaching out. According to our confidentiality document, customer data submitted for translation to Azure Cognitive Services Translator (both standard and custom models), Speech service, the Microsoft Translator Speech API, and the text translation features in Microsoft Office products are not written to persistent storage. There will be no record of the submitted text or voice, or any portion thereof, in any Microsoft data center. The audio and text will not be used for training purposes either. The documents you upload using Custom Translator (portal or APIs) are stored encrypted in your workspace. Custom Translator uses your uploaded documents exclusively to provide your personalized translation system and does not use it for any other purpose. The documents you upload to Custom Translator will be stored in the Azure region you selected when you created your Translator key until you delete them or until your account expires. For more details on Microsoft Translator and how to use the service, please review our documentation.

Hope this helps!