Oct. 23, 2019, 9:56 a.m.
MBIE has announced a $13 million investment in Te Hiku Media for Papa Reo, a multilingual language platform that will develop cutting edge natural language processing tools, starting with te reo Māori. The tools will enable applications to be built that will ensure all New Zealanders can use te reo Māori when engaging with their digital devices. New Zealand English and Pacific languages will also be included to further support the growth of a multilingual Aotearoa.
Formed in 1991, Te Hiku Media is an organisation committed to the revitalisation of tikanga and te reo Māori. Founded as a not-for-profit, Te Hiku Media serves as a monument to the demonstrated courage and action of the community during the critical Māori rights recognition period. Kaumātua (tribal elders) established Te Hiku Media as an innovative iwi broadcaster to share local topical issues in te reo Māori, gather stories and maintain te reo o te kāinga (language of the home) as a method of communication. In 2013, kaumātua from Ngāti Kuri, Te Aupōuri, Ngāi Takoto, Te Rarawa and Ngāti Kahu encouraged the CEO, Peter-Lucas Jones, to pursue technological innovation and secure a digital future for te reo Māori in a rapidly moving and uncertain digital landscape.
Peter-Lucas Jones says, "Te Hiku Media will continue to play a critical role in Māori language revitalisation and iwi broadcasting by improving and maintaining native Māori language pronunciation and expression through digital innovation and Māori data sovereignty. Te Hiku Media is an experienced kaitiaki of Māori data and iwi led corpus gathering campaigns for the promotion and growth of te reo Māori and tikanga Māori."
"Te Hiku Media are pleased to receive this funding and lead the development of the data science platform, Papa Reo. This award celebrates the efforts of the haukāinga to innovate digitally and it recognises the work we have done over the last six years to build digital capability in the science and te reo Māori spaces."
Te Hiku Media gained a Vision Matauranga Capability Fund placement in 2015 with Engineer and Physicist Keoni Mahelona to build science and innovation capabilities within the organisation. Then from 2017/2018, Te Hiku Media were Ka Hao: Māori Digital Technology Fund recipients and developed the first te reo Māori automatic speech recogniser in a project called Kōrero Māori. Te Mātāwai also funded Te Hiku Media for the development of a digital pronunciation prototype for te reo Māori.
"The data science platform funding is the culmination of years of hard work and proving that Te Hiku Media can deliver groundbreaking digital innovation from the haukāinga. Crucial to this has been and will continue to be our strategic relationships and dynamic collaborations, and we look forward to developing our newest strategic project relationship with Dragonfly Data Science."
Peter Lucas Jones says that Wellington-based Dragonfly Data Science was engaged during the Kōrero Māori project to support Te Hiku Media with the development of speech tools for te reo Māori. Dragonfly specialises in the application of machine learning and statistical analysis, and their expertise in data science makes them a key partner for Papa Reo.
Alongside New Zealand-based academics and experts in linguistics, te reo Māori and machine learning, the investment will allow for the collaboration to extend beyond Aotearoa, bringing in critical global expertise including Cambridge University’s Speech Group and Oxford University academics. A unique collaboration for the project is with the machine learning group at Mozilla, a non-profit tech company known for the Firefox web browser.
Papa Reo will build Aotearoa’s data science capability in machine learning and data stewardship through kaitiakitanga. The project team includes relationships with academic institutions who will provide supervision for PhD students and post-doctoral fellowships. We look forward to hearing from people interested in machine learning, artificial intelligence and te reo Māori.
For all media enquiries contact:
027 368 9241