Google Translate now offers translations with reduced gender bias

By Avatar on
December 10, 2018
  -   Like us now!  
 

In order to reduce gender bias, Google has updated Google Translate to offer translations in both masculine or feminine form. Previously, the translation was only given in one gender. At the moment, this feature is limited to a few languages with support for more languages to come in future. Product manager, James Kuczmarski, said in a blog post, “Over the course of this year, there’s been an effort across Google to promote fairness and reduce bias in machine learning.”

Gender bias has existed in Google Translate because of how Google has trained their algorithm. The data used to train the algorithm can be a big reason as well. So it would associate words like strong as masculine attribute and beauty as feminine. “So when the model produced one translation, it inadvertently replicated gender biases that already existed,” said Kaczmarski. “For example, it would skew masculine for words like ‘strong’ or ‘doctor,’ and feminine for other words, like ‘nurse’ or ‘beautiful.’”

At the moment Google Translate supports reduced gender bias for English to Spanish, French, Italian, Portuguese and Turkish to English. Google plans on adding support for more languages soon. It also plans to cater to situations involving non-binary genders. At the moment it is limited to the website but Google says that they have plans of bringing it to iOS and Android. They also plan to address gender bias in autocomplete.

The move comes after the company came under criticism last month for offering gender bias pronouns in Gmail. Earlier this year, Google was also criticized for offering gender bias translations, but it looks like Google is trying to fix the situation. The problem is complex and it will take some time before Google can truly eliminate gender bias.

Like our stories? Follow our Instagram for pictorial updates.Follow @techjuicepk

 
Student from Peshawar develops electric bike which can go 102 km on a single charge
 
 
 
Microsoft's facial recognition technology will abide by these six principles