What does Google say about Minor Tweaks To Machine Translations?

Google recently released a blog post around tweaks they allow when it comes to machine translations. The blog gave examples and some good advice. In the past, though, there have been issues where Google allowed minor tweaks to machine translations, resulting in some user experience problems. 

At Google’s October SEO office hours, the question was raised about whether it is acceptable to use automatically translated content that has been reviewed by a human and subjected to only minor editing changes. 

Google SEO-Office Hours Episode

In this episode of Office Hours, we follow a new format where questions are submitted in writing, and the Google team subsequently gives answers. 

Lizzi Sassman (@okaylizzi) is a tech writer who “cares for” the Google Site Central documentation and joins us today alongside John Mueller.

The new format of the live show doesn’t allow for any follow-up questions from the audience, so the answers tend to be very similar to what you would find in Google’s documentation. This can be a problem if you’re looking for clarification on an issue or concern.

For example, the person asking the question above was worried about the content that had been translated from another language using a machine. They employed human editors to review the content regularly, and it was usually found to be acceptable, with only minor changes needed.

The person asking the question is concerned about whether “minor tweaks” are enough to make the content acceptable for Google. Lizzi Sassman gave an answer that closely followed Google’s documentation.

The question explicitly asks if minor tweaks are good enough for Google. After all, Google is a constantly evolving platform, and even the smallest changes can have a big impact. The answer to that question depends on the quality of the translated content. If the quality is good, then minor tweaks may be all that is needed. However, if the quality is poor, more significant changes may be required.

Does Google Accept Moderately Edited Machine Translated Content?

Users asked:

“A site uses machine translation to offer posts in other languages.

The content is reviewed by human translators and they’re often happy with the quality after the minor tweaks.

Is this okay for Google?”

Google’s Lizzi Sassman answered:

“Well, that’s good to hear that the human translators are happy and this is totally fine for Google as long as there’s a human involved in the review process. That’s the key.

The thing you want to watch out for is making sure that the quality continues to be good and working well for the humans that are reading the content.”

The answer doesn’t specifically mention if minor edits are okay, only that if the “human translators” are happy with it, then Google should be happy too.

Could it be that Google doesn’t check if the content is machine-translated but relies on standard content quality signals? We’re not sure. The new Office-Hours format doesn’t allow for follow-up questions, so we can’t ask for clarification.

Google Spam Policies

Automated text translation tools are generally considered to be spammy, but there are exceptions when a human element is involved. This is according to Google’s developer documentation about spammy content.

This is what Google’s documentation says:

“Examples of spammy auto-generated content include:

Text translated by an automated tool without human review or curation before publishing”

So, it’s clear from Google’s published guidelines that as long as a human is editing the machine-translated content, it will be okay with Google.

Additionally, in a Google Office-Hours video from April 2022, John Mueller mentioned how AI-generated content can sometimes be considered spam. He said how auto-translated content can sometimes be an issue and spoke about AI content-generating tools and compared them to auto-translation tools. He said that while AI-generated content is not yet perfect, it is getting better all the time and will eventually be able to produce high-quality content on a variety of topics.

At the 24:55 minute mark of the April 2022 Office-Hours video, Mueller stated:

“I think, I don’t know, over time, maybe this is something that will evolve, in that it will become more of a tool for people.

Kind of like you would use machine translation as a basis for creating a translated version of a website.

But you still… essentially work through it manually.”

Why Should You Check Auto-Translated Content?

As we mentioned above, Google’s main concern is that the content users find when they search using the engine results pages (SERPs) is high quality and that they will be happy with what they find.

However, something that wasn’t discussed is that translated content often contains tell-tale signs that a machine translation algorithm can identify.

Many years of research have been dedicated to detecting machine-translated content.

A research paper published in 2021 found that machine-translated text can be complex for humans to detect. The study, titled “Machine Translated Text Detection Through Text Similarity with Round-Trip Translation,” examined how well humans can identify machine-translated text when compared to the original version.

Besides, out of 100 translated texts, human raters could only identify the source language of just over half of the texts.

The researchers noted:

“The average accuracy was 53.3% (55.0% for the native speakers and 52.0% for the non-native speakers), which was close to random.”

The Text Similarity With Round-Trip Translation (TSRT) approach outperforms human raters and scored higher than state-of-the-art translation detectors when the paper was published in 2021.

This approach can help to determine the native language of any given translated text. Moreover, it can also determine which translation algorithm did the translation.

They reported:

“The evaluation results show that TSRT outperforms other methods, with an accuracy of up to 90.2%.

Moreover, TSRT could also identify the original translator and translation language with 93.3% and 85.6% of accuracy, respectively.”

Although it’s unclear if Google can detect translated content, we know that technology exists that can detect it better than humans. This technology can determine which translation algorithm did the translation.

If the fact that it’s against Webmaster Guidelines and may have a negative user experience isn’t enough to motivate you to edit machine-translated content, then perhaps the possibility that Google is analyzing content quality for machine translation might be a reason to give that kind of content a more thorough review.

Conclusion

You can publish machine-translated content if it’s been edited for quality. If you find that the quality of the machine-translated text is not suitable for your website, you should edit the text yourself, or have someone else edit it, for quality. It’s clear from Google’s published guidelines that as long as a human is editing the machine-translated content, Google will be happy with it. So, when making these edits, you need to ensure that the edited text reads naturally and doesn’t seem like it’s been translated by a machine.

If you’re reading this, we hope you enjoyed our latest blog! And if you have any other questions or concerns about Minor Tweaks to Machine Translations, please contact us anytime at EverRanks.