article · Discover Data
The migration of academic talent creates significant challenges for national development and knowledge systems in developing countries such as Ghana. Bibliometric records from Web of Science and Scopus between 2007 and 2020 provide a longitudinal dataset tracking disciplines, institutional connections, and mobility patterns. Applying data mining and association rule learning helps uncover discipline-destination movement patterns, using metrics such as support, confidence, lift, and conviction to evaluate predictive strength. In addition, time series modelling forecasts future movement across various academic fields. The findings indicate that agricultural and biological sciences, health sciences, and social sciences face the highest risk of sustained talent loss over the long term. These predictive analytics offer a reproducible methodology to support data-informed decision-making for national research systems seeking to improve talent retention, institutional capacity, and diaspora engagement.
The loss of highly qualified academics threatens the sustainability of national knowledge systems and hinders socio-economic progress in emerging economies. Using data-driven forecasting helps policymakers identify vulnerable academic sectors before brain drain causes critical deficits. This enables research institutions and governments to direct resources towards targeted talent retention initiatives, institutional development, and diaspora collaboration programmes where they are most needed.
The methodology offers an applied predictive analytics framework that could be adapted into policy-planning software or human resource decision-support tools for national science councils, ministries of higher education, and university administrators. While currently an analytical model tested on retrospective bibliometric data, it could inform future institutional analytics platforms aimed at monitoring workforce retention and managing research talent across emerging higher education sectors.
AI-generated from the published abstract. Always read the original work before citing.
International organizations, policymakers and researchers have taken notice of the many facets of international mobility and its impact on both the countries of origin and destination. International mobility of research talent has become a critical concern for a developing nation like Ghana, where the departure of highly educated people, particularly in academia, poses significant barriers to national prosperity and the sustainability of the knowledge system. This study employs data on research publications and machine learning techniques to develop a robust system for assessing and forecasting the migration tendencies of Ghanaian researchers to other countries. A longitudinal dataset comprising research disciplines, connections, and mobility trends was created by extracting bibliometric data from Web of Science and Scopus covering the years 2007–2020. Using metrics like support, confidence, lift, and conviction to assess the strength and predictability of emigration rules, the study uses data mining techniques to generate association rules of the field of disciplines to uncover important discipline-destination migration patterns. Time series modeling was used to estimate future migration trends across academic disciplines, revealing that the most vulnerable fields to long-term talent loss are the social sciences, agricultural and biological sciences, and health sciences. The study’s conclusions demonstrate the urgency of data-driven policy initiatives targeted at diaspora participation, institutional building, and talent retention. The suggested paradigm improves knowledge of academic mobility in the Ghanaian context and offers a reproducible strategy for other countries handling the transfer of research talent. The study emphasizes how crucial predictive analytics are to national research systems’ strategy human resource management decision-making.
This page summarises published work. The authoritative version sits with the publisher.
DOI: 10.1007/s44248-026-00110-5
Is something wrong with this record? Report it or request removal.
Discussion
Have you built on this work, tried to replicate it, or seen it applied in practice? Share what you know. Verified researchers and MARATTO™ domain experts can open a discussion, and any member can reply. Contributions are reviewed before they appear.
No discussion yet. Open the first thread.
New to MARATTO™? Create a free account.
