Developing Interactive, Educational Material on Deep Learning

The impact of Artificial Intelligence in our society has grown immensely. Providing proper education on this topic is therefore crucial. The goal of our research is to promote AI literacy by making Deep Learning conceptually understandable to the younger generation. This study will propose a solution by discussing and evaluating our developed course material addressing the research questions: How to explain pattern recognition? How to explain the basics of machine learning? How to explain the basics of deep learning? We conducted a pilot for upper pre-university secondary education students to evaluate the performance of the developed course based on its effectiveness, coherence, and usability. The data was collected using tests, surveys, and evaluation discussions. We found that our course can successfully explain pattern recognition and sufficiently explain the majority of the basics of both ML and DL algorithms with interactive material. Some material did raise confusion on the details of the classification model but was successful for the remaining learning goals. These findings are significant as they show a promising method to foster a conceptual understanding of deep learning integrated into the standard school curricula. Whereas current educational tools often focus on interacting with AI, our course actively uncovers the inner mechanisms explaining why AI works.