Bambara-French Machine Translation Competition


Welcome to the Orange Silicon Valley Bambara-French Machine Translation Competition. This competition focuses on the development of a machine translation model for Bambara, a low-resource language, as part of an overall goal of improving global language representation across technologies.

Bambara is a one of the closely related Manding languages, which are part of the larger Mandé language family. It is predominantly spoken in Mali and is one of the country’s 13 national languages, with 15 million speakers and 5 million native speakers.

Low-resource language challenges

Recent advancements in Neural Machine Translation have made the development and design of high-performing translation models for high-resource languages an accessible task for big tech companies with access to large sets of parallel, translated data.

However, low-resource languages — ones that do not have extensive digital data sets and online communities, often because they are primarily spoken and not written — benefit much less from these recent advancements.  Additionally, linguistic research is often lacking in low-resource languages, making conventional natural language processing tasks more difficult.

Meet Allahsera Auguste Tapo

The competition

In this competition you will create Bambara-to-French and French-to-Bambara translation models. See our evaluation and competition details page and our rules page for additional specifics.

To register for this competition you must fill out the form at the bottom of this page. After we receive your email, you will be given full access to our training data.

Bambara-French translation competition registration form:



Sarah Luger
Principal, Technology Group

Please contact Dr. Sarah Luger with any questions about the Bambara-French Machine Translation Competition.

Hosted by Orange Silicon Valley

Orange Silicon Valley is a US subsidiary of Orange, one of the world’s leading telecommunications operators serving more than 253 million customers across 27 countries. Learn more about us here.