A Data-Driven Framework for Holistic Student Performance Evaluation and Industry Readiness Using Machine Learning
Year: 2025
Type: article
Abstract: The growing disparity between higher education outcomes and rapidly evolving industry requirements highlights the urgent need for an integrated, data-driven evaluation approach. Traditional assessment models focus primarily on academic performance, neglecting critical employability skills such as communication, adaptability, and problem-solving. This study introduces a novel multi-tier evaluation framework leveraging machine learning (ML) techniques to holistically assess student performance and predict industry readiness. Using a decade-long dataset (2015–2025) comprising 750 undergraduate students from Tamil Nadu, India, the proposed system integrates cognitive, behavioural, and technical skill indicators. Decision tree classifiers, particularly the J48 model, are utilized for predictive modelling, supported by clustering algorithms for deeper analysis. The model achieved a notable accuracy of 86%, significantly outperforming traditional evaluation methods. The findings demonstrate how AI-driven predictive analytics can bridge the gap between academia and industry by enabling timely interventions and personalized development plans. This research contributes a scalable, explainable, and context-sensitive solution for workforce preparedness in the Indian higher education landscape.
Authors: Vijayakumar Krishnan, Dr. Naveen A
Institution: Thiruvalluvar University
Language: English
Subfield: Computer Science Applications
Field: Computer Science
Domain: Physical Sciences
Open Access status: bronze
OpenAlex ID: W4414492056