Digital Literacy Education: A Deep Analysis Through Machine Learning and Education Research

Abstract

AI and digital literacy education are interrelated, where AI helps define the set parameters of the digital universe, and digital literacy empowers learners to understand it. Digital literacy trains individuals to critically evaluate AI-generated content, identify biases, and leverage AI-enhanced tools ethically. Implementing AI in education promotes being informed for both consumers and producers in an AI-based society. This study investigates the application of machine learning (ML) models in assessing and predicting digital literacy skills across different educational contexts. Utilizing a large dataset, the accuracy of five regression models was calculated: Random Forest, Ridge, Lasso, Support Vector Regression (SVR), and K-Nearest Neighbors (KNN) based on Mean Squared Error (MSE) and R <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> values. Results showed linear techniques (SVR, Ridge, Lasso) worked almost perfectly (MSE ≈ 0.00, R <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> ≈ 1.0) in comparison to distance and ensemble metrics. The Random Forest showed little overfitting, whereas greater errors by KNN (MSE: 6.31-7.59) illustrated sensitivity to feature scaling. These findings support the idea that predictors are linearly dependent on each other (e.g., pre-training outcomes, educational qualification) and outcome measures for digital literacy. The research fills gaps in vocational and adult learning studies, offering actionable recommendations to address digital divides through targeted interventions.

Ähnliche Arbeiten