In today’s education system, it is important to have fast and efficient algorithms for categorizing and generating exam questions across cognitive levels. Automating this process aids in assessments that align with learning objectives, improving both evaluation and learning outcomes. Bloom’s taxonomy, a framework that classifies learning objectives into different levels of complexity and understanding, is widely used in modern pedagogy by promoting higher-order learning and critical thinking. This paper presents a comprehensive review of existing approaches for both question classification and automated question generation, making it the first study to analyze these two aspects together. It focuses on methods ranging from traditional machine learning algorithms to advanced deep learning techniques, evaluating their effectiveness in improving efficiency and accuracy across both domains. With the growing interest in taxonomy-based learning approaches, this research is very useful in offering insights into how AI-driven techniques can improve question classification and automated question generation. The main finding of this review is identifying research gaps, particularly the absence of practical application of these methods in real-world educational tools. There is increasing interest in designing syllabus based on Bloom’s taxonomy. By bridging this gap, future research can focus on developing intelligent educational platforms that dynamically generate and classify questions to enhance student learning. The aim of this review is to provide an overview of recent approaches and techniques proposed in question classification and automated question generation. It provides valuable insights for researchers and educators aiming to implement efficient and scalable techniques in real-world educational environments, eventually improving the design and implementation of question generation and question classification tools.