COBE is designed as a brainstorm mediator. To recognize the voices of the meeting participants a supervised learning algorithm was used that could classify audio data into prelabelled categories. A mostly functional prototype was created, integrating the algorithm. A lo-fi prototype was used to conceptually prove the concept and the model was proven to work with an accuracy of 99.79%. This design demonstrates an implementation of XAI in an educational context. COBE supports users in their brainstorming process through an embodied visualization of contribution by the successful implementation of the SMO learning algorithm.
Course                   Embodying Intelligent Behaviour in Social Context (DPM120), Q1, 2020-2021
Grade                     8
Tools                       Weka, Raspberry PI, Python, Lo-fi Prototyping
Team members     Chantal Vriens, Daphne Vermeer, Katja van Weert, Maud Oomen

You may also like

Back to Top