arXiv subjects

Oche D. Ankeli

Publications and source records attributed to Oche D. Ankeli.

2 recordsLinked to original sources

arXiv2025

IyaCare: An Integrated AI-IoT-Blockchain Platform for Maternal Health in Resource-Constrained Settings

Maternal mortality in Sub-Saharan Africa remains critically high, accounting for 70% of global deaths despite representing only 17% of the world population. Current digital health interventions typically deploy artificial intelligence (AI), Internet of Things (IoT), and blockchain technologies in isolation, missing synergistic opportunities for transformative healthcare delivery. This paper presents IyaCare, a proof-of-concept integrated platform that combines predictive risk assessment, continuous vital sign monitoring, and secure health records management specifically designed for resource-constrained settings. We developed a web-based system with Next.js frontend, Firebase backend, Ethereum blockchain architecture, and XGBoost AI models trained on maternal health datasets. Our feasibility study demonstrates 85.2% accuracy in high-risk pregnancy prediction and validates blockchain data integrity, with key innovations including offline-first functionality and SMS-based communication for community health workers. While limitations include reliance on synthetic validation data and simulated healthcare environments, results confirm the technical feasibility and potential impact of converged digital health solutions. This work contributes a replicable architectural model for integrated maternal health platforms in low-resource settings, advancing progress toward SDG 3.1 targets.

cs.CY↗

arXiv2025

AfriStereo: A Culturally Grounded Dataset for Evaluating Stereotypical Bias in Large Language Models

Existing AI bias evaluation benchmarks largely reflect Western perspectives, leaving African contexts underrepresented and enabling harmful stereotypes in applications across various domains. To address this gap, we introduce AfriStereo, the first open-source African stereotype dataset and evaluation framework grounded in local socio-cultural contexts. Through community engaged efforts across Senegal, Kenya, and Nigeria, we collected 1,163 stereotypes spanning gender, ethnicity, religion, age, and profession. Using few-shot prompting with human-in-the-loop validation, we augmented the dataset to over 5,000 stereotype-antistereotype pairs. Entries were validated through semantic clustering and manual annotation by culturally informed reviewers. Preliminary evaluation of language models reveals that nine of eleven models exhibit statistically significant bias, with Bias Preference Ratios (BPR) ranging from 0.63 to 0.78 (p <= 0.05), indicating systematic preferences for stereotypes over antistereotypes, particularly across age, profession, and gender dimensions. Domain-specific models appeared to show weaker bias in our setup, suggesting task-specific training may mitigate some associations. Looking ahead, AfriStereo opens pathways for future research on culturally grounded bias evaluation and mitigation, offering key methodologies for the AI community on building more equitable, context-aware, and globally inclusive NLP technologies.

cs.CL↗