Improving the Usability of Intelligent Interface Designs

Abstract

Rapid advances in artificial intelligence (AI) have led to its widespread integration into various aspects of daily life. To create user-centred experiences, it is critical that AI methodologies incorporate usability principles to ensure effective, efficient and satisfying interactions for users. However, challenges arise from the complexity of real-world scenarios and the limitations of established design practices, requiring careful evaluation during system development. This study aims to investigate potential evaluation criteria for the usability assessment of AI-based software systems. The methodology involves synthesising findings from relevant research, defining characteristics of derived principles, condensing information into guidelines, and validating their relevance and structure. By synthesising these efforts, a set of initial guidelines is presented to improve the understanding of how to enhance usability in AI-integrated systems. The proposed considerations aim to establish a more human-centred relationship with AI systems by personalising outcomes while respecting privacy and configuration preferences, with the goal of improving usability and natural interaction between humans and AI.

Keywords: Artificial Intelligence, Usability Evaluation, User Interface Design, Guidelines, Natural Interaction

DOI: 10.54941/ahfe1005436