#low resource languages

[allthingslinguistic]

The many languages missing from the internet

English and a handful of other languages dominate the internet, but this is leaving indigenous cultures without a voice online. Now they are fighting back.

An article in BBC Future about linguistic representation on the internet. 

The Kaqchikel Mayan community from Guatemala includes more than half a million speakers. Miguel Ángel Oxlaj Kumez is part of it and was one of the organisers of the first Latin American Festival of Indigenous Languages on the Internet, held in 2019.

“When I get on the internet I find more than 90% of the content in English and hence a significant percentage in Spanish and other languages,” he says. “So what I have to do is to move to another language, and that favours the displacement of my own language.”

“It discredits my own language, because – as it is not on the internet – then it is not valid, then it does not work, therefore why am I going to continue learning it? Why am I going to teach it to my children if, when I turn on the internet or television, I cannot find it there?”

Oxlaj Kumez is working with other activists to create a version of Wikipedia in Kaqchikel Mayan, as well as a translated version of Mozilla’s Firefox web browser. His dream is to be able to have a “digital life in my own language, and when I decide to move to another language that it will be my decision”.

He is not the only one with that dream. In 2003, Unesco adopted a recommendation to promote the use of multilingualism online. Ever since, the organisation has been pushing for universality on the internet, with a special focus on indigenous languages.

Read the whole thing. 

[allthingslinguistic]

The poetic process powering real-time language translation in Namibia

A series of intimate conversations could teach an AI to understand both language and culture.

An interesting article about the social process of creating machine translation datasets in Khoekhoegowab. Excerpt: 

On the surface, Wilhelmina Ndapewa Onyothi Nekoto and Elfriede Gowases seem like a mismatched pair. Nekoto is a 26-year-old data scientist. Gowases is a retired English teacher in her late 60s. Nekoto, who used to play rugby in Namibia’s national league, stands about a head taller than Gowases, who is short and slight. Like nearly half of Namibians, Nekoto speaks Oshiwambo, while Gowases is one of the country’s roughly 200,000 native speakers of Khoekhoegowab.

But the women grew close over a series of working visits starting last October. At Gowases’s home, they translated sentences from Khoekhoegowab to English. Each sentence pair became another entry in a budding database of translations, which Nekoto hopes will one day power AI tools that can automatically translate between Namibia’s languages, bolstering communication and commerce within the country.

“If we can design applications that are able to translate what we’re saying in real time,” Nekoto says, “then that’s one step closer toward economic [development].” That’s one of the goals of the Masakhane project, which organizes natural language processing researchers like Nekoto to work on low-resource African languages.

Compiling a dataset to train an AI model is often a dry, technical task. But Nekoto’s self-driven project, rooted in hours of close conversation with Gowases, is anything but. Each datapoint contains fragments of cultural knowledge preserved in the stories, songs, and recipes that Gowases has translated. This information is as crucial for the success of a machine translation algorithm as the grammar and syntax embedded in the training data.

Read the whole thing.