Self-Adaptive User Interface for Elderly Users: A Bayesian Network Approach

Using modern and advanced applications can be challenging for older individuals particularly with faint vision. This paper introduces a real-time self-adaptive interface developed specifically for elderly users, conquering challenges like unclear vision and memory loss using Bayesian algorithms. Unlike many existing solutions, which tend to be either very complex or rigid to specific models, the proposed Bayesian algorithm performs accurately in modeling uncertainty, employing probabilistic calculations to dynamically shape the UI based on user behavior. The main goal is to enhance accessibility by detailing the system concept, technological approach, and user needs elicitation process. While the current implementation only focuses on dynamically adjusting text and icon size based on user behavior, including misclicks and feedback, the paper further presents a plug-and-play, generic model for the unobtrusive assessment of cognitive load and vision-related challenges.