Is Google a mouth piece for China?

Hi Jimmy and other users, 

 

Welcome to the user moderated Google Translate Support Community. 

 

I don't understand the Chinese language but I believe you. I think its appropriate to give you some context regarding how such things happen. Just so you know Google Translate allows users to contribute to the languages available in the Google Translate Community. Users could also install the Google Crowdsource app and help Google improve the translations. Looks like the wrong translation got approved by the Google Translate algorithm here. 

 

Given the popularity of Google products, some users with a with a malicious intent abuse the crowdsourcing functionality. These users, sometimes become successful in getting their word across and in the process brings bad press to Google as well.  

 

Users in the past have abused Google Search, Google Maps and many a times vandalized Wikipedia articles too. 

 

I have tested this " So sad to see Hong Kongbecome China " and in the languages I am familiar with  ( Arabic, Urdu, Hindi, Telugu, and Kannada ) and the translation is correct. The issue is only with the Chinese translation. 

With regards,

W16 engine

Arabic and Urdu --

 

Hi there, 

Overall, of course it's great to fix the issue and we're grateful for the explanation, but I think it would help to clarify a few things if possible:

> Translate is an automatic translator, using patterns from millions of existing translations to help decide on the best translation for you.

This is a pretty general statement of how the software works in general. For this specific case, was the problematic translation based on 'patterns from millions of existing translations'? Was it always translated incorrectly for Hong Kong, Taiwan, etc.? Or was it a recent change? If it was a recent change, this seems inconsistent with the explanation.

> These automatic systems can sometimes make unintentional mistakes like translating a negative to a positive. We appreciate the feedback and we are working on improving the technology.

This is also a general statement of how the system works in general. In this specific case, did the system make the mistake of translating a negative to a positive without intent (intent not necessarily from Google, but from others)?

The concern that people have is that this offensive translation seemed to apply to Hong Kong, Taiwan, and other frequently targeted groups specifically. People are concerned that it was not a broadly applied mistake, but a specifically targeted manipulation of the product to affect specifically targeted groups of people (perhaps victims). Was that the case or not?

What are the details of exactly what happened in this case?

@MosheTLV thanks for the note, I think that makes sense. However, I'm not sure specifics of their algorithms are needed here. For example, we could have some response like, "In this case there was an issue with an algorithm" or "In this case acts from an outside party caused an issue with an algorithm and Google is working to prevent similar incidents in the future." Would that be general enough while still addressing the issue at hand?

That doesn't seem much more specific than what was posted, which generally provides information that the algorithm uses patterns from existing translations:
> Translate is an automatic translator, using patterns from millions of existing translations to help decide on the best translation for you.

Has Google never provided details about any specific problem in the past (while not divulging specifics of an algorithm)?

I was able to do these tests when the issues happens

To start a sentence with "so sad/ so unhappy/ so upset" is a literal translation of Cantonese into English 

This is a "fun" translation locals use to mock at each other of their "poor" English thus this is very rare to appear in normal day to day normal English communications

All of "so sad/ so unhappy/ so upset" gets incorrect translation that morning and of now

I am concern there might be other undiscovered cases out there. (Case 4A and 4B.jpg)

These results shows the incorrect translation is specifically targeting HK, hong kong and Tibet

Reply above from google staff and gold expert suggest the error is un-intentional then

1. Why is this problem only lasted for around 3 hours, is it really that easy to get fixed for such short time?

2. Why this problem only happen to the rare sentence structure of "so XX to see Hong Kong becomes China"?

 I do feel there is some "optimization" done for this to happen

Please throw some facts to explain and stop trying to explain without fact

It just make me feel I cannot put my trust on google.

Tried to attached 4-6 image in new post but got error shall retry attach again latter

Appreciate suggestion, but with facts please.Regards,

Brian

I found that :

所以遗憾地看到香港成为中国

translated to:

So I regret to see Hong Kong become China. [confused implication, awkward grammar]

which I clarified to:

I so regret seeing Hong Kong become China. [polite, sympathetic]
I very much regret seeing Hong Kong become China. [emphatic]
I regret seeing Hong Kong become China.  [laconic]

I entered this into Google Translate:

臺灣藜並非我國農政單位表定之糧食作物，使政府在推動種植及產品開發上受到相當大的限制

and get this:

Taiwanese glutinous rice is not a food crop designated by China's agricultural administrative units, which has caused the government to impose considerable restrictions on promoting planting and product development.

---

Ignoring the "rice" issue (the article is about 紅藜 / red quinoa, not rice of any kind), why does it mention China? 我國 = "our country's" - here a reference to Taiwan's national institutions, not China's.