Rinn Artificial Intelligence at MTU

Rinn Artificial Intelligence is a national research and innovation centre designed to advance foundational research in Data Science and AI while supporting open, international, interdisciplinary and transdisciplinary research addressing major societal challenges.

Rinn Artificial Intelligence – Research & Innovation in Data Science and AI is a national research and innovation centre supported by Taighde Éireann – Research Ireland. It brings together researchers across foundational and applied AI, Data Science, health, governance, society and culture.

Programme facts verified via Research Ireland (retrieved 2026-07-18). Rinn AI is directed nationally by Prof. Noel O'Connor (DCU); Dr Afli's role is institutional (MTU), not national.

Dr Haithem Afli's involvement

Dr Haithem Afli serves as Institutional Co-Lead for Rinn Artificial Intelligence at MTU, Deputy Theme Lead for Inclusive Language Model & Translation Methods, and Principal Investigator.

Inclusive Language Model & Translation Methods

The Inclusive Language Model & Translation Methods theme focuses on language technologies that work effectively across languages, cultures, communities and domains. It addresses the limitations of current language models and translation systems, particularly for under-resourced languages, culturally specific knowledge and communities that remain poorly represented in mainstream AI development.

Multilingual and under-resourced language models

• language models for Arabic and other under-resourced languages
• cross-lingual transfer
• multilingual representation learning
• efficient adaptation of large language models
• linguistic inclusion beyond high-resource English-centric systems

Inclusive machine translation

• translation for low-resource languages
• culturally sensitive translation
• preservation of meaning, sentiment and pragmatic context
• multimodal and domain-specific translation
• human-centred translation evaluation

Cultural reasoning and representation

• culturally grounded reasoning
• cross-cultural differences in language-model behaviour
• cultural knowledge representation
• evaluation of bias, stereotyping and cultural misalignment
• Global South perspectives in AI development

Reliable evaluation of language models

• limitations of LLM-as-a-judge evaluation
• benchmark validity
• human and expert evaluation
• evidence hierarchies for AI assessment
• robustness, uncertainty and reproducibility
• evaluation across languages and cultures

Responsible and human-centred AI

• transparent and explainable language technologies
• participatory evaluation
• accessibility and linguistic fairness
• social and ethical effects of language technologies
• AI systems designed around human and community needs

Translation methods for specialist domains

• healthcare and biomedical translation
• scientific communication
• public-sector communication
• financial and institutional language technologies
• educational and digital-literacy applications

Research contribution at MTU

At MTU, Dr Afli's Rinn AI activity will connect foundational language-model research with existing strengths in Human-Centred AI, multilingual NLP, responsible evaluation, computational biology, healthcare AI and research commercialisation. The work will support collaboration across computer science, language research, social sciences, health, industry and public-sector partners.

Rinn AI, ADAPT and the HAI Research Group

These three structures are complementary but not interchangeable: Rinn AI and ADAPT are distinct national centres, and the HAI Research Group is Dr Afli's own MTU group. Their funding and governance are separate.