Connect with us

Africa

Meta’s AI Machine Translation Research Helps Break Language Barriers

Published

on

 Meta s AI Machine Translation Research Helps Break Language Barriers
Meta’s AI Machine Translation Research Helps Break Language Barriers

1 Today, Meta (www.Facebook.com) announced that it has created and open sourced ‘No Language Left Behind’ NLLB-200 (https://bit.ly/3yJo7KS), a unique AI model that is the first to be translated 200 different languages ​​including 55 African languages ​​with state of the art results. Meta is using the modeling techniques and learnings from the project to improve and extend translations on Facebook, Instagram, and Wikipedia.

nigeriannewspaper

2

nigeriannewspaper

3 In an effort to develop high-quality machine translation capabilities for the world’s most low-resource languages ​​(https://bit.ly/3Av3FyB), this unique AI model was designed with a focus on African languages. They are challenging from a machine translation perspective. AI models require lots and lots of data to help them learn, and there isn’t much human-translated training data for these languages. For example, there are more than 20 million people who speak and write Luganda, but it is extremely difficult to find examples of this written language on the Internet.

nigeriannewspaper

4 We work with professional translators for each of these languages ​​to develop a reliable benchmark that can automatically assess translation quality for many low-resource languages. We also work with professional translators to do human evaluation, which means native speakers of the languages ​​evaluate what the AI ​​produced. The reality is that a handful of languages ​​dominate the web, so only a fraction of the world can access content and contribute to the web in their own language. We want to change this by creating more inclusive machine translation systems, unlocking access to the web for the more than 4 billion people around the world who are currently excluded because they don’t speak one of the few languages ​​in which it is available. the content.

5 “It is impressive how much AI is improving all of our services. We have just opened up an AI model that we believe can be translated into 200 different languages, many of which are not supported by current translation systems. We call this project No Language Left Behind, and the AI ​​modeling techniques we use are helping to make high-quality translations for languages ​​spoken by billions of people around the world. To give an idea of ​​scale, the 200-language model has over 50 billion parameters, and we train it using our new Research SuperCluster, which is one of the fastest AI supercomputers in the world. Advances here will enable more than 25 billion translations every day in our apps. Cross-language communication is one of the superpowers that AI brings, but as we continue our AI work, we improve everything we do, from displaying the most engaging content on Facebook and Instagram to recommending more relevant ads and maintaining our services. safe for everyone. ”, Meta CEO Mark Zuckerberg said in a post on his Facebook profile (https://bit.ly/3nJEBwo).

6 Language is our culture, identity and lifeline with the world. However, because high-quality translation tools do not exist for hundreds of languages, billions of people today are unable to access digital content or fully participate in online conversations and communities in their preferred or native languages. This is especially true for the hundreds of millions of people who speak the many languages ​​of Africa.

7 “Africa is a continent with a very high linguistic diversity and language barriers exist every day. We are pleased to announce that 55 African languages ​​will be included in this machine translation research, making it a breakthrough for our continent,” said Balkissa Ide Siddo, Director of Public Policy for Africa, speaking about the launch of the AI ​​model. . “In the future, imagine visiting your favorite Facebook group, finding a post in Igbo or Luganda, and being able to understand it in your own language with the click of a button; that’s where we hope an investigation like this will take us. Highly accurate translations in more languages ​​could also help detect harmful content and misinformation, protect the integrity of elections, and curb cases of sexual exploitation and human trafficking online.”

8 Commenting on accessibility and inclusion in the quest to build an equitable metaverse, Ide Siddo added: “In Meta, we are working today to ensure that as many people as possible can access the new educational, social, and economic opportunities that the Next the evolution of the Internet will bring the technology of the future and an experience of everyday life tomorrow”.

9 To confirm that the translations are of high quality, Meta also created a new evaluation dataset, FLORES-200 (https://bit.ly/3aqMnrM), and measured the performance of the NLLB-200 in each language. The results revealed that NLLB-200 outperforms the prior art by an average of 44 percent.

10 Meta is also opening up the NLLB-200 model and publishing a host of research tools to enable other researchers to extend this work to more languages ​​and build more inclusive technologies. Meta AI also provides up to $200,000 in grants to nonprofit organizations for real-world applications for NLLB-200.

11 There are versions of Wikipedia in more than 300 languages, but most have far fewer articles than the more than 6 million available in English. Following Meta’s partnership with the Wikimedia Foundation (https://WikimediaFoundation.org), the non-profit organization that hosts Wikipedia and other free knowledge projects, modeling techniques and learnings from NLLB research are now also shared. apply to the translation systems used by Wikipedia. editors Using the Wikimedia Foundation’s content translation tool (https://bit.ly/3NNCMcf), articles can now be easily translated into more than 20 low-resource languages ​​(those without extensive datasets to train AI systems), including 10 that were previously not supported by any machine translation tools on the platform.

12 To explore a demo of NLLB-200 showing how the model can translate stories from around the world, visit here (https://bit.ly/3AxuouI). You can also read the research paper here (https://bit.ly/3AASuEV).

13

shop bet bet9ja

NNN is a Nigerian online news portal that publishes breaking news in Nigeria, and across the world. We are honest, fair, accurate, thorough and courageous in gathering, reporting and interpreting news in the best interest of the public, because truth is the cornerstone of journalism and we strive diligently to ascertain the truth in every news report. Contact: editor @ nnn.ng. Disclaimer.