The increasing importance of computational literacy drives efforts to introduce computer science concepts early in education. However, many primary school teachers lack formal technical training, making selecting engaging, pedagogically sound activities difficult. Existing tools often lack personalization, leading to ineffective choices, while disparities in technology access highlight the need to integrate unplugged, digital, and hybrid approaches.
This thesis proposes a recommendation system to help teachers select suitable computer science activities based on age group, format, resources, and educational objectives. The system will leverage a comprehensive tagging framework grounded in pedagogical principles and compare different recommendation approaches for accuracy and scalability. This research aims to empower teachers with a personalized tool that boosts confidence in teaching computer science while enhancing student motivation and creating more inclusive learning experiences.