Language acquisition assessment involves evaluating ŕuency, intonation, coherence, vocabulary, conjugation, and grammar. The learning of national languages, which is the support of the local culture, is nowadays a priority for several African governments. In Cameroon, with over 200 national languages from different linguistic families, implementing this assessment in schools is challenging due to limited teacher capacity and varying student native languages as well as reading deőciencies. With the lack of adequate tools for early assessment and remediation of basic reading skills, it is impossible to determine whether a student’s poor academic performance is due to a lack of decoding skills, language proőciency, or insufficient knowledge to answer questions. To address this, an automatic speech assessment solution leveraging Natural Language Processing (NLP) could assess students in their mother tongue, enhancing training efficiency. AI enhances natural language processing by providing techniques and tools to leverage text and speech in many applications. In this paper the development of AI models for the assessment of learning African languages in primary education in African countries (particularly in Cameroon). We are proposing guidelines to deal with the linguist, the lack of guidelines for the learning assessment and a few data challenges of these languages. A generic new framework of skills assessment based on the level of learning task-related deőciency is also proposed to overcome the limit of the classical assessment approaches. The trained AI models for automatic speech recognition will be integrated in the implementation of this assessment approach.

This publication focuses on the following countries.

CM

Cameroon

Filter by country

The following resources are related to this publication by citation.