Project course "Data Science and Artificial Intelligence" in the classroom
Finding a free parking space via app - that is the task of the students Maarten Liphardt, Karl Schneidewind and Luis Wengenmair of the twelfth grade (Q2) of the Gymnasium Theodorianum in Paderborn. In the course "Data Science and Artificial Intelligence", they developed their own artificial intelligence (AI) supported web application to predict the occupancy of Paderborn parking spaces and car parks (parkpb.maar10media.de). The course was held for the third time as part of the "Data Science and Big Data in Schools" (ProDaBi) project at the University of Paderborn. The ProDaBi project was initiated by the Deutsche Telekom Foundation as well as financially supported since 2019 and is led by Prof. Dr. Rolf Biehler, Didactics of Mathematics, and Prof. Dr. Carsten Schulte, Didactics of Computer Science, both from the University of Paderborn. One aim of the project is to develop a curriculum on data science and artificial intelligence in upper secondary schools.
Through the project course, knowledge in the areas of statistics, data collection, data analysis, machine learning, decision trees (directed rule system for decision making) and artificial neural networks is to be imparted. With their newly acquired knowledge, the students were now able to develop their own data-driven web application to predict free parking spaces. This allowed them to demonstrate how artificial intelligence programmes can be used for the practical benefit of society. The twelve years of parking space data processed with an AI model were provided by the companies responsible for parking space management in Paderborn, ASP from Paderborn and RTB from Bad Lippspringe.
Putting what they have learned into practice
During the course, the students learned about different tools and programming and scripting languages in order to use them in the development of the web application and the underlying AI models. They followed a self-directed learning and working process. They organised their teamwork within the framework of agile project management or an agile learning process according to the so-called Scrum method. They were supported by the project course staff Annika Löhr (Gymnasium Theodorianum Paderborn), Yannik Fleischer (Mathematics Didactics, University of Paderborn) and Sven Hüsing (Computer Science Didactics, University of Paderborn).
The students' tasks included cleaning, evaluating and visualising data as well as configuring, training and applying data-driven machine learning models to predict parking space occupancy. They also documented and presented the data processing and modelling in the form of so-called computational essays, which provide insight into the process of data analysis and the development of artificial intelligence as well as the corresponding results. Finally, they integrated the results of the machine learning model into their own web application.
Last month, the students of the Theodorianum presented the success of their work at an online product presentation. They showed the audience their AI-supported web app and presented more detailed information about the data set used, the structure of the underlying AI models and the design of the user interface. When asked what the students liked best about the project, they replied: "It was particularly exciting to see what is hidden behind artificial intelligence, which you otherwise only come into contact with superficially. Particularly interesting were the questions: how do you create an artificial intelligence yourself, how elaborate is it and how can you comprehend it?"
The ProDaBi project is still looking for more cooperation partners to test the project course concept and the teaching materials. Interested schools in the Paderborn area are welcome to contact prodabi(at)campus.upb(dot)de.
Website of the project: www.prodabi.de
