Building collaborative resources for poorly endowed languages community-oriented modeling

The applications of natural language processing (NLP) today feed a large part of Indo-European languages, with a large body of quality data available in large quantities. As open source, data corpora in African languages are almost non-existent, how can the advances in NLP be secured for these poorly endowed languages ? In this article, we address the problem of constructing lexicographic resources.
We wish to introduce a model for building lexical resources by exploiting the socio-linguistic skills of local linguistic communities. Throughout the sections, we will present the new dictionary coding model resulting from this community-oriented modeling.