The fast-changing pace of the African workspace has been evident in several institutions as more women have risen through the rank-and-file meritoriously within the past decade. One of the easiest ways a society measures her progress is through academic institutions. This research uses Nigeria’s academia as a cohort to examine how much female visibility has changed over the years, using academic research output as a yardstick. This research intends to enquire about the number of published peer-reviewed articles in Nigeria attributed to female researchers. The authors’ names were obtained from the Web of Science using the address of the first author as the inclusion criterion. Preprocessing was performed to remove duplicates and correct spelling inconsistencies. We built two machine learning models (logistic regression and sequential LSTM) to predict the gender of the authors. After tokenization and vectorization at the character level, we performed sequential training on 70% of the dataset with logistic regression as the baseline model, followed by the LSTM model. Model performance was evaluated on test data using accuracy, precision, recall, and F1-score. The sequential LSTM model outperformed logistic regression on all metrics, indicating that it has a great capacity for learning complex representations of data through multiple layers. The results also indicate a steady rise in women's contributions to research, implying progress toward gender equality in Nigerian academia. This study provides insights into the strengths and limitations of each model for gender prediction, highlighting their value in uncovering societal and institutional trends based on gendered data.

Authors:
Promise Gift Philip, University of L'Aquila, Italy
Chibuikem Chrysogonus Nwagwu, SINTEF Manufacturing, Norway
Ogonna Angela Ndubuisi, University of L'Aquila, Italy

About the Presenter(s)
Philip Promise Gift is currently a masters student of the University of L'Aquila, L'Aquila, Italy. His interest is data science, big data, machine learning, and artificial intelligence. His current project is Assessing Gender Dynamics in Nigerian Academic Research. A Data-Driven Approach Using Machine Learning

Connect on Linkedin
https://www.linkedin.com/in/promise-philip-a7a3b51b9/

See this presentation on the full schedule – Thursday Schedule

Bookmark