XR Management Training Simulator supported by Content-Based scenario recommendation

Extended Reality (XR) is actively bringing innovative opportunities to business, from Virtual Reality (VR) assistants to Augmented Reality (AR) sales advertisements. However, there is currently a lack of adaptive profiling in training simulators, limiting their potential, effectiveness, dynamicity, and customization. This paper proposes the use of a Content-Based Recommender System integrated into an XR simulator to predict training scenarios for the user. Although customization can be achieved in many ways, this project focuses on in-game input data: user interaction with virtual objects, decisions made by the user, and metadata associated with the training scenario (plot keywords, characters involved in the scenarios, objects metadata, scenario category and room in which the scenario takes place). The results of the user study show that the participants enjoyed the simulation and preferred a dynamic and customized training narrative compared to a standard linear one. This innovative approach will enable in the future the automation of personalized training simulations, creating targeted and individual XR experiences for each user.