Can Google Identify Translated Content?

Koloktronis

Registered Member
Joined
Dec 29, 2015
Messages
67
Reaction score
44
I am bilingual and I operate in a niche that has tons of content in my native language.

I take the content in my native language, I translate it, edit it, pass it through a plagiarism checker and I republish it to my English speaking blog.

I am wondering if "G" can understand that the content I am translating/republishing is translated.

If yes, how?

If not, do you know any news or technology that is about to be launched that will be able to identify translated content?
 
Ask yourself: How can Google detect AI generated content which is technically, random by default?
Nowadays with AI generated translations is difficult to find two identical translations of the same content, still there are many elements that are equally shared.

Google can clearly identify plagiarized content as long as is less than 1-5% of plagiarism in the text (in fact this has been easy for ages, more than 10 years). As long as Google can use translation tools to contrast your content (specially their own neural network) you will be spotted.

But trust me that there is a ton of people doing massive sites with thousands of translated content pages, ranking better than you and me with zillions of organic visits per day.

So, here is my rule of thumb:

If you are planning to go super-wide translate, use AI whatever you like. Maybe some of the pages within will be spotted while others will not, but the key here is the massive automation in the process.
But if you are planning to build less than a hundred pages, then try to, at least, to rewrite the translated content to make it look very unique.
 
had no problem with translated content sofar, i don't think google is checking for that.
also imagine google would have to check each and every page and translate it to 1000s of languages and cross compare.
even with their computing power that would take ages.
 
I translated so far way more than 500,000 Pages into different languages.
In my test, as long as the translation is not directly from or to English, everything is fine. My content is ranking as hell.

For example: English -> French -> your desired source language.
 
I would say if i would translate the content and re-write with a spinbot and u got ur own unique content
 
First off, you'd have to spend a good deal of time proofreading those articles
Otherwise, the bounce rate will kill those pages
A better alternative would be to use GPT-3 based AI writers
 
I translated so far way more than 500,000 Pages into different languages.
In my test, as long as the translation is not directly from or to English, everything is fine. My content is ranking as hell.

For example: English -> French -> your desired source language.

500,000 pages... Impressive..

So you are saying that if....

Italian >> English = Problem

Italian >> French >> English = No Problem?
 
I would say if i would translate the content and re-write with a spinbot and u got ur own unique content
Every translated article from [Native Language] >> [English] is being checked via plagiarism checkers before published.

Are you suggesting that this is not enough?
 
Google is all powerful my friend. Nowhere to hide from Him. :)
Nah.. I don't think they're that powerful.. Don't underestimate human ingenuity

Also, always keep in mind that where's an action there a reaction as well...

If "G" takes action, the Black Hat community will react..
 
No, Google does not view translated content as duplicate, as content in another language is different than if it is appearing in the same language twice. However, it is recommended to have a human translation as Googlebots can tell if the site has been translated using Google Translate and may view it as 'spammy'.
 
Personally, I translate everything myself and haven't had any issues yet. It is natural to restructure sentences when writing in a different language. Something "G" translate can only do to some extent.
 
Personally, I translate everything myself and haven't had any issues yet. It is natural to restructure sentences when writing in a different language. Something "G" translate can only do to some extent.
Thank you for your input.

By the way, I don't use "G" translate. I used a paid AI translating tool.
 
No, Google does not view translated content as duplicate, as content in another language is different than if it is appearing in the same language twice. However, it is recommended to have a human translation as Googlebots can tell if the site has been translated using Google Translate and may view it as 'spammy'.
I use a paid translation service - not G translate. The translation offered by the paid translation service is stunning. Way better than G's.

Do you think G bots will be able to "spot" them?
 
At the current point in time, as it pertains to the SERPs/G Updates, I'd say no.

But, you have to know that's coming eventually. More so, they will eventually "penalize" AI content sites with:

- False information.
- Poor/bad/ grammer.
- Broken/mixed sentences.


I address this in my latest black SEO for beginner's thread.
 
At the current point in time, as it pertains to the SERPs/G Updates, I'd say no.

But, you have to know that's coming eventually. More so, they will eventually "penalize" AI content sites with:

- False information.
- Poor/bad/ grammer.
- Broken/mixed sentences.


I address this in my latest black SEO for beginner's thread.

I think so to..

But what if your editorial team (or you) proof checks all information and content produced by AI.

For example:

- All information published is verified.
- No grammatical or spelling errors
- Content actually gets edited. No copy/paste from the AI tool or translation tool as is.

Do you still websites that take time to ensure the content generated via translation or AI tools is accurate, will still get penalized?
 
I think so to..

But what if your editorial team (or you) proof checks all information and content produced by AI.

For example:

- All information published is verified.
- No grammatical or spelling errors
- Content actually gets edited. No copy/paste from the AI tool or translation tool as is.

Do you still websites that take time to ensure the content generated via translation or AI tools is accurate, will still get penalized?

I have multiple websites right now that are fully AI generated content. I actually have a journey thread about it that I need to update.

I believe that if you:

Proofread for factual inaccuracies, grammatical errors, and sentence fragments, that there's no way a GPT-3 form of detection software could give out anything more than a false positive.
 
Italian >> English = Problem

Italian >> French >> English = No Problem?
I tested January 2020 this behavior with 10 blogs. Each blog had 1000 pages.
6 Months later, June 2020:

Italian >> English => 30% Pages ranked
Italian >> French >> English => 80% Pages ranked

All translations are human readable, without grammatical errors. Since my test, I'm using my 80% method without any changes and it works really well.
 
Usually the problem with translated content is getting reported by the author.

I found some shitbag using my content in a different language. I told him either give credit or take it down. I got hundreds of links back and let him use the content in his native language.
 
Back
Top