CV
Basics
| Name | Abiodun Modupe |
| Label | Lecturer & Coordinator/Research Scientist |
| abiodunmodupe@gmail.com | |
| Phone | +1 (832) 774-3303 |
| Url | https://dupsys.github.io |
| Summary | Dr. Abiodun Modupe earned his Ph.D. from the University of the Witwatersrand, Johannesburg, South Africa, and conducts research at the intersection of Artificial Intelligence (including natural language processing), Computer Vision, Cybersecurity, and Health Informatics. His work focuses on designing data-driven and vision-enabled intelligent systems, leveraging large language models and big data methodologies to generate impactful insights and optimize system performance. His research applications span digital security, healthcare technologies, and environmental monitoring. |
Work
-
2019.11 - Present Lecturer and Coordinator/Researcher Scientist
University of Pretoria, Pretoria. South African
Conduct research at the intersection of AI, with a focus on NLP, computer vision, and cybersecurity, and have 6+ years of experience providing technical leadership in machine learning and data infrastructure. I have a proven track record of managing million-dollar AI research initiatives and leading or collaborating with teams of researchers to develop innovative ML systems for NLP challenges.
- Led curriculum design, supervised research projects, and served as Big Data degree coordinator.
- Published multiple papers in top-tier NLP conferences and journals.
- Developed multiple funding proposals for research projects.
Awards
- 2013.08.27
Best Student Paper Award
International Conference of Information Security and Internet Engineering (IAENG 2013)
The paper titled: Investigating Topic Models for Mobile Short Messaging Service Communication Filtering.
Publications
-
2025 Mafoko: Structuring and Building Open Multilingual Terminologies for South African NLP
Proceedings of the Sixth Workshop on Resources for African Indigenous Languages (RAIL), Co-located with DHASA2025
The study introduces Mafoko, a RAG-based framework that consolidates South Africa’s fragmented terminology into open datasets, improving translation accuracy and advancing equitable multilingual NLP technologies that better showcase the country’s linguistic diversity.
-
2025 Cross-lingual embedding methods and applications: A systematic review for low-resourced scenarios
Natural Language Processing Journal
This study presents a systematic review of cross-lingual transfer learning (CLTL) techniques, focusing on their foundations, applications, evaluation methods, language coverage, and recent advances, and highlighting future research directions for low-resource NLP.
Languages
| English | |
| Native speaker |
| Yoruba | |
| Native speaker |