Pioneering Somali-language
AI through world-class research & Transformative Technology

“Peer reviewers emphasised that our work on fake news and toxicity detection makes a significant contribution to combating misinformation and toxicity on Somali social media, addressing critical gaps in online safety for low-resource African languages.”

“Reviewers noted that the SomaBERTa model and its accompanying datasets establish an important benchmark for Somali-language AI, enabling researchers and practitioners to build more advanced models and expand NLP applications in the region.”

“Multiple reviewers described the research as a strong and impactful contribution to African NLP, with well-supported results demonstrating clear improvements over existing multilingual and African language models.”

“Reviewers also recognised our research on Somali lemmatization corpus development as a foundational contribution to Somali-language NLP, providing the first large-scale, linguistically grounded lemmatization lexicon and annotation platform that fills a critical resource gap for low-resource African languages.”

“Peer reviewers highlighted that the expert-validated lexicon and web-based annotation platform create a sustainable and expandable infrastructure for Somali and other morphologically rich African languages, supporting long-term inclusive AI development.”

“Reviewers emphasised that this resource will strengthen downstream NLP tasks, from POS tagging to information retrieval and hybrid neural-symbolic systems, positioning Somali as a well-supported language in the next generation of African AI technologies.”