Combining intelligent tutoring systems and automatic story generation for personalized tutoring with mobile robots Type : Master Thesis

R. Sumner

Abstract:In the last years, we have been teaching programming to children using the Thymio mobile robot (bottom, left) and a visual programming environment (bottom, right). This action has proven successful in schools where children are supported by professional educators. However, at home, children are likely to be left alone facing programming exercises, feel lost, and give up. Recent research results, in learning technologies and video games, provide evidence that artificial intelligence techniques can be used to improve this situation. The practical goal of this project is to build a tutoring app that provides an engaging experience for children learning programming with the Thymio robot. The scientific goal is to assemble an intelligent tutoring engine that combines a probabilistic approach and adaptive algorithms for automatically generating the story and the levels. The app should propose a sequence of challenges to the user, in which each challenge trains a certain programming knowledge. To maintain the user in a feeling of flow, the content of the challenges should be adapted in function of her/his progress. At every step, a method called ZPDES can select the knowledge that the learning progress of the user. Then, methods from procedural content creation in video games should build an interesting challenge and integrate it into a consistent story. The result should form a tutorial that provides an engaging and non-repetitive adventure game. This project will be implemented in Qt-QML and C++, with some minor contributions to the script of the robot. The student should combine recent methods from the state of the art in novel ways and adapt them to the context of the project.

Education,Computer Science