Interaction Design for AI Systems: An oriented state-of-the-art
Firmino Silva,André Costa
DOI: https://doi.org/10.1109/HORA55278.2022.9800084
2022-06-09
Abstract:In recent years, new web-based technologies have emerged, and mobile devices and applications access have become widespread, resulting in new paradigms in Human-Computer Interaction (HCI). These paradigms created several challenges in interaction design regarding customization, adaptability, and accessibility. Over the lifetime of a digital product, the user must adapt to updates, changes in the interface, and new functions, while having to provide attention to customization and configuration tasks in interfaces created for millions of individuals worldwide. Simultaneously, users' conditions change, and the system must adjust to new requirements, preferences, and needs in a fast-paced digital environment. And new functions, interfaces, and technologies that meet current market directions, such as Artificial Intelligence, Augmented Reality, Virtual Reality, and even Design without physical interfaces must be developed and implemented. This research work explores the importance of Interaction Design (IxD) in the present and future Artificial Intelligence (AI) systems, highlighting the prominent authors of User-Centered Design and usability principles, guidelines, and heuristics. Moreover, researching appropriate design principles for developing a positive, effective, and safe interaction between humans and computers. This study will also deepen and orient the study of the state-of-the-art to promote the exploration of HCI in a period of enormous challenges and opportunities in AI. As a result of this research work, a table summarizes, compares, and classifies the state of the art according to Usability & Design Principles. Outputting a matrix that aggregates the fundamental principles for designing interfaces in artificial intelligence systems, regardless of their interface. This matrix serves as a base for a framework to create a prototype, in future work, based on the guidelines suggested.
Computer Science