A scientometric analysis and systematic review of scientific literature on the validation of computer science graduate employability factors and predictive models

Abstract

Promoting student employability stands as a central objective for educational institutions, often serving as a barometer of their effectiveness. However, the landscape of the job market is undergoing rapid transformation, driven by forces such as globalization, automation, and the rise of artificial intelligence (AI). Understanding the pivotal determinants of employability and constructing predictive models can yield substantial benefits for all stakeholders. This knowledge empowers students to make more enlightened career choices by discerning their strengths and areas that require development. In this study, we use scientometric analysis and a systematic literature review (SLR) to delve into recent trends and future trajectories within the realm of identifying, validating, and constructing predictive models for employability factors about computer science (CS) graduates. Our research encompasses 592 pertinent studies published between 2010 and 2023, sourced from Scopus, a pivotal academic database. Through keyword co-occurrence and author co-citation analyses using VOSviewer and CiteSpace software, we scrutinize network parameters and vital data. This review primarily strives to chart the progression of research within the field of employability, pinpoint knowledge gaps, and chart a course for future investigations. The scientometric analysis uncovers four notable clusters in the cited articles, encompassing subjects such as factors influencing the employability of CS graduates, models for validating these factors, predictive models for employability, and the impact of employability matrices. Our SLR offers invaluable insights into the prevailing validation and predictive models for employability among CS graduates. Guided by our SLR, we propose that forthcoming research should explore the potential of innovative AI techniques to pinpoint key factors and elevate the precision of predictive models geared toward computer science graduates' employability.