Evaluating teachers’ effectiveness in classrooms: an ML-based assessment portfolio

Effective teachers are strongly committed to creating a positive learning experience, instinct, and impact for transforming students’ learning. The transforming elements can be defined through related attributes, such as self-efficacy, regular attendance, and cooperative behavior. However, this involves a significant data analysis task to measure teachers’ performance and predict their effectiveness in the education domain. Underpinned by a recognized design perspective of design science research, this study establishes a methodological framework for designing a solution artifact utilizing machine learning algorithms informing the design science research for IS design. We designed a new solution artifact utilizing a case dataset, specifically, a record of the UCI machine learning repository. Researchers can measure teachers’ effectiveness through the proposed innovative technique that elevates distinct resources to configure learning opportunities and relevant monitoring of learning. To evaluate the proposed ML model in measuring teachers’ effectiveness, we validated the prediction by contrasting it with other comparable models. We developed two ML models using K-means and hierarchical algorithms and found that the K-means presented the best outcome in representing three clusters: negative, positive, and neutral feedback, also showing 99% accuracy using the random forest classifier. Therefore, the K-means clustering technique is selected to be the core component of the solution for predicting teachers’ effectiveness.