Over in Aotearoa New Zealand, a Maori-led media company is using Artificial Intelligence to preserve the Maori Language and protect Indigenous data sovereignty
Te Aupōuri, Ngāi Takoto, Ngāti Kahu man Peter-Lucas Jones is the CEO of Te Hiku Media.
He’s helped establish Papa Reo, a Maori language transcription application designed to keep Te Reo Maori and other Indigenous languages alive.
Mr Jones told NIRS News a goal of the program is to remove the English language’s impact on the Maori language.
“We believe that there’s evolution that is natural language evolution, but there’s also white assimilation.
And decolonising the sound of our language allowed us to recognise the influence of English on pronunciation.”
Mr Jones says the idea of the program stems from the principles of Maori-led radio stations.
We are one of 21 Iwi radio stations or tribal radio stations.
And the reason that we exist is because of language loss and language decline.
Maori language is the medium for transmission of all things Maori.
But for many generations, the language was literally beaten out of our parents and grandparents.
So we serve to promote Maori language and Maori culture through broadcasting.
And as a platform, we enable the members of our community to share the voices on issues, both topical and current, that are dear to the hearts of the people that are members of our tribal groups.
As a tribal radio station, and by virtue of our origins, we’re an organisation that is committed to the revitalization of our language and our culture
And that’s when we decided to speak native sound into our future..”
In order to create a program with an 86 per cent success rate Te Hikiu needed to use it’s extensive back catalogue to build it’s data set, and to fill in any gaps volunteers were ready to help.
Elders in particular helped generate what is called a “corpus’ – a collection of spoken material that can be read and interpreted by machines.”
“That corpus could be about what is the name of that home? How did that hill get that name? What is the name of that river, and what is the name of the water that flows in it?
So the nomenclature around our landscape plays an important part in our corpus gathering activities.” Mr Jones says.
Te Hiku hopes other First Nation groups around the world get inspired by their work and create their own data science projects.
Mr Jones also highlighted why it’s important for these initiatives to be led by Indigenous groups.
“Data has land.
And we must always be in control of our data.
We know what happened with the land.
And the thing about the online world, if we don’t have a place that we can call our own, we will be landless in that place too.”