How do datasets, developers, and models affect biases in a low-resourced language?

Re-assign community