Books
Peer‑Reviewed Articles
Each article is directly associated with its foundational literature to demonstrate deep mastery of the field.
Article 1 (Flagship) – Empirical, Neurodiversity & AI
Acoustic Signatures of Neurodiversity: Using ASR Error Cascades as Computational Biomarkers in Low-Resource Classrooms
Targets: IJAIED, Computers & Education, or LAK (Learning Analytics and Knowledge)
Language: English (Technical & Medical impact)
Timeline: September 2026 (co-authored with Prof. Marcelo Worsley)
Key References: Radford et al. (2022) Whisper; Worsley & Blikstein (2018) Multimodal Learning Analytics; Yeung & Alwan (2018) ASR for Children; Lyon et al. (2003) Definition of Dyslexia; Ntsama (2022) Doctoral Thesis.
Article 2 – Theoretical Formalization & Computational Modeling
From Philosophical Framework to Predictive Model: Formalizing the ‘Cognitive Trapeze’ as a Bayesian Network
Targets: Minds and Machines, British Journal of Educational Technology
Language: English
Timeline: M8 (Month 8 of MSCA project)
Key References: Vygotski (1997) Thought and Language; Piaget (1975) Equilibration; Varela et al. (1991) Embodied Mind; Hutchins (1995) Cognition in the Wild; Spikol & Cukurova (2020) Multimodal Learning Analytics; Chomsky (1968) Language and Mind.
Article 3 – Pedagogical Impact & Global South
Evaluating the Impact of an Offline AI Tutor on Semantic Comprehension: A Controlled Pilot Study in Cameroonian ZEPs
Targets: International Journal of Educational Development, Learning and Instruction
Language: English
Timeline: M16 & M22 (Empirical validation)
Key References: World Bank & UIS (2022) Learning Poverty; UNICEF & ITU (2020) Internet Access at Home; UNESCO (2023) GEM Report; Holmes & Miao (2023) Generative AI in Education; Ntsama (2023) Web Social.
Article 4 (Priority Technical) – Benchmarking Quantized ASR-LLM Pipelines
Benchmarking Quantized ASR-LLM Pipelines (Whisper Large + Qwen 0.5B) for Educational Feedback in Low-Resource, Offline African Classrooms
Preprint Deposit: arXiv (early deposit for community feedback)
Final Targets: IJAIED, IEEE Transactions on Learning Technologies
Language: English
Timeline: Preprint on arXiv (December 2026) → Final submission (M22)
Key References: Radford et al. (2022) Whisper; Jacob et al. (2018) Quantization; Worsley (tiilt lab) Equity in Learning Technologies; Ntsama & Worsley (2026) Preprint arXiv.
Article 5 – Ethics, Policy & Decoloniality
Decolonizing AI in Education: Why Offline, Neurodiversity-Aware Tutors are the Key to Bridging the Digital Divide
Targets: AI & Society, Globalisation, Societies and Education
Language: English
Timeline: Open (Post-M22)
Key References: Mohamed et al. (2020) Decolonial AI; Selwyn (2016) Is Technology Good for Education?; Ollivier & Puren (2011) Web 2.0; Ntsama (2023) Web Social.
All research outputs are deposited in open-access repositories with permanent DOIs, in full compliance with FAIR principles.
Published Datasets (4):
corpus_reference.csv, wer_evaluations.csv| License: CC-BY 4.0 | Indexed in: OpenAIRE |
| License: CC-BY 4.0 | Indexed in: OpenAIRE |
corpus_reference.csv, learner_productions.csv, generated images, README