Customisable Voice Interaction: Feature Model Configuration by Means of Deep Learning
Kasassov, Clara-Sophie, Jorde, Manuel, Lischinski, Eric. Customisable Voice Interaction: Feature Model Configuration by Means of Deep Learning. Complex Practical Task. 2022.
The aim of our project work within the scope of the complex practical training was to apply software variability to implement customisable language interaction, with a deep learning approach chosen to configure feature models. Thus, a feature model should be created for the language interaction that includes all user-specific customizations of the conversation and later implements a deep learning algorithm that returns the correct configuration for individual users. The end result should be a configuration mechanism that can adapt the language interaction to various user characteristics such as disabilities, interests, or language usage, thereby personalizing the dialogue individually.