Design and implementation the data oriented aspects of an innovative framework that supports educational smartphone games

OData support
Supervisor:
Dr. Forstner Bertalan
Department of Automation and Applied Informatics

At the Department of Automation and Applied Informatics we have been developing a framework that supports innovative educational mobile games for years. Our purpose is to make the process of learning for children (primarily with learning disabilities) more effective and enjoyable. To achieve this goal, on the one hand the AdaptEd project puts the learning into a digital environment known and loved by the current generation by using tablets, and on the other hand it uses several commercially available sensors (e.g. EEG, EKG, eye tracking device) to calibrate the joint educational game to run with optimal settings (difficulty, task type, etc.).

The data from the operation of the framework (sensor data, events occurring during the game, results, screenshots, etc.) are valuable in multiple respects; if they are properly visualized then the special teacher can use them to get an overall picture of the game’s course, the child’s performance and state without close proximity, or later in time. In addition, we can examine the process and the effectiveness of learning with their analysis. However, it is important in both aspects that we possess proper and distortion-free data. We need to be able to filter out those real life cases when the children cheat, e.g. they have their friends or parents do the exercises.

During my work, I planned and implemented the collection of data, and their storage both on the client and the server side. I created a user-friendly web interface for the teachers and the developers of the framework to manage the entities connected to the framework (users, supervisors, institutes, games, etc.), and to display the gameplays in a tabular and a graphical way. I analysed the possible visualisation techniques of the data with multiple technologies. Finally, as the first data analysis task I created an algorithm that can detect suspicious gameplays that may be frauds. In order to do that I applied multiple approaches, classification algorithms and methods to the problem, tested and evaluated them. The data collected and cleaned such way can be used for later analysis.

Downloads

Please sign in to download the files of this thesis.