While we still continue to improve the Swahili to English MT system, we
also work on the English to Swahili MT system. Also in it, we use
rule-based methods as far as possible, and the use of statistical
methods is used for solving such semantic ambiguity, where reliable
rules cannot be written.
text-to-text MT, we plan to boost the speech recognition
and speech production research on Swahili. This will be done
in cooperation with the research teams that have excelled in