AutoBloom: AI-Based Classification of Cognitive Levels in Educational Content

Amidst the dynamic landscape of modern education, the effective use of classroom recordings remains a challenge. Particularly, the manual analysis of the cognitive depth and breadth of teaching materials can be dauntingly time-consuming and potentially biased. In response to this challenge, This research proposed AutoBloom, an AI-based framework that automates the classification of teaching materials according to Bloom's Taxonomy. By leveraging deep learning models to analyze multimodal content in online classroom recordings, AutoBloom accurately assigns attributes of Bloom's Taxonomy to instructional materials. This automation provides educators with valuable insights into the cognitive levels targeted by their teaching materials, enabling data-driven decisions in curriculum design and teaching methodologies. AutoBloom not only addresses the need for efficient and objective analysis of educational content but also offers scalability and accuracy. The framework incorporates the concept of Shapley values to enhance explainability, ensuring transparency in the classification process. Additionally, AutoBloom facilitates continuous evaluation and improvement of teaching approaches, serves as a benchmarking tool for comparing strategies and content, showcases the potential of AI in analyzing multimodal data, and addresses ethical considerations related to data privacy and security. The findings highlight the transformative impact of AI in providing deeper insights for educators and institutions to enhance teaching effectiveness and improve learning outcomes in today's rapidly evolving educational landscape.