This is the dataset used in the research paper: https://lanfrica.com/record/a-machine-learning-based-text-classification-and-machine-translator-for-selected-under-resourced-languages Contains 1,820 lines of translations.