Classifiers’ competency in identifying digital innovation skills for teachers of Bisha Province, Saudi Arabia

Developing teachers’ digital innovation skills secures community education and advances the rank of the national school system. This paper assesses teachers’ digital innovation skills (DISs) in Bisha Province, Saudi Arabia, and examines the associated challenges. The dataset used consists of a substantial sample of 400 local teachers from this area. Several evaluation methods were implemented, but the machine learning (ML) classifier was the main research methodology used/examined to identify DIS problems, to predict educational entertainment settings, and to help stakeholders design an appropriate DIS training module for local educators. Powerful models were examined by calculating key classification performance criteria, such as accuracy, recall, precision, F1-score, confusion matrix, average SHAP values, and AUC curves. The extreme gradient boosting (XGB) classifier was the top-ranking algorithm, and it exhibited the highest performance scores after optimizing both scalar and visual classification outputs. The revealed results were significant and identified the teachers’ digital innovation status by introducing key features of measurement. The findings suggested positive implications for pedagogy strategies and teachers’ development. The recommendations of this paper have the potential to support stakeholders in advancing teachers’ innovation.