Evaluating Intelligibility Usage and Usefulness in a Context-Aware Application

Abstract

Intelligibility has been proposed to help end-users understand context-aware applications with their complex inference and implicit sensing. Usable explanations can be generated and designed to improve user understanding. However, will users be willing to use these intelligibility features? How much intelligibility will they use, and will this be sufficient to effectively improve their understanding? We present a quasi-field experiment of how participants used the intelligibility features of a fully-functional intelligible context-aware application. We investigated how many explanations they willingly viewed, and how that affected their understanding of the application's behavior, and suggestions they had for improving its behavior. We discuss what constitutes successful intelligibility usage, & provide recommendations for designing intelligibility to promote its effective use.

Conversations with my washing machine

“…They clearly wanted to evaluate their convenience against the benefits of shifting the load. This finding echoes existing research on intelligibility of context-aware application [19].…”

Designing human-in-the-loop autonomous Cyber-Physical Systems

“…Works of Lee et al [36] and Stanton et al [37] encourage the development of systems that keep the human informed, provide him/her with feedback and reduce cognitive responsibilities and stress. Related to these works, other ones that also provide insights on this issue are the work of Nahavandi in [38], which explores trust as a main attribute of systems to be accepted in our daily lives, and the work of Lim et al in [39], which discusses intelligibility in context-aware applications.…”

Have You Met Your METs? – Enhancing Patient Motivation to Achieve Physical Activity Targets in Cardiac Tele-rehabilitation

“…An intelligible system allows users to understand and even learn about the inner working of the system. It has been shown that intelligible systems are easier to use and are more trusted by their users (Lim and Dey, 2013). In order to add credibility to systems that are designed to support rehabilitation, it becomes even more essential for patients to be able to trust the system.…”