Skip to content
Projects

Language

Ejagham Language Model

A language model for Ejagham, a low-resource language of Nigeria and Cameroon, to preserve the language and give its speakers working text tools.

Animation of John 11:17 in Ejagham, with its tone marks. Verse text from the project's Ejagham New Testament data.

The language

Ejagham (ISO 639-3 etu) is an Ekoid language spoken on both sides of the Nigeria–Cameroon border.

Ejagham is written with the letters ə, ʉ and ŋ, plus tone marks. Tone carries grammatical meaning, such as aspect and mood.

The current work

We build translation models from the New Testament alone. We test the models on whole books held out from training.

Native speakers are now judging the translations.