Predicting Mentoring Effectiveness in a Computer Science Program: A Machine Learning Approach

Abstract

Mentoring is a critical academic tool to positively influence students' outcomes. While there is a broad consensus about the benefits of mentoring, still there is a divergence of opinion regarding the attributes based on which student mentees evaluate the effectiveness of a mentoring program. This study has drawn a sample from undergraduate students of a computer science program and applied educational data mining to predict mentoring effectiveness. WEKA machine learning linear regression technique was applied to a primary dataset. It was observed that academic subject knowledge support was considered the most important predictor of mentoring effectiveness.