AI Based Assessment: Theoretical Foundations, Specific Models, and the Directions of Classroom Application

This study aims to explore why AI-based assessment is necessary, what types and criteria it involves, and what specific examples and future directions for development and application exist in the growing field of AI-based assessment tools. To achieve this, the study establishes two key research questions: first, what are the necessity, core principles, types of AI-based assessment, and the criteria for evaluating its effectiveness, and second, what are the latest technological features applied to AI-based assessment tools, and how can these features be reflected in specific tools and classroom applications? The findings of this study are summarized as follows: first, effective utilization of AI-based assessments can yield several educational benefits, including reduced teacher workload, increased focus on teaching activities, and enhanced opportunities for interaction with students; second, to ensure the validity and fairness of AI-based assessment tools, it is necessary to consider factors such as reliability, validity, fairness, usability, data security, and the applicability and flexibility of the evaluation; third, the study presents specific examples of AI-based assessment categorized by four major principles and features, including personalization and adaptability, continuous diagnosis and feedback, multidimensional assessment, and data-driven decision-making; and fourth, recent technological trends in AI-based assessment include the datafication of teaching and learning experiences using natural language processing, comprehensive collection and analysis of assessment materials through AI-based portfolio assessment, evaluation of activities within the metaverse via virtual space and simulation learning assessment, and analysis of the impact of emotional factors on learning through sentiment analysis for learning attitude assessment. This study concludes that the integration of generative AI technologies into the field of school education assessment offers significant implications for enhancing assessment practices.