Explore a 21-minute conference talk that delves into user-reported problems with intelligent everyday applications. Analyze the findings from a study of 35,448 reviews of popular apps like Facebook, Netflix, and Google Maps, uncovering issues related to algorithmic decision-making. Learn about the sentiment analysis and topic modeling techniques used to identify key problem areas, including content, algorithm functionality, user choice, and feedback. Discover insights from a follow-up survey revealing users' coping and support strategies when facing these challenges. Gain valuable implications for designing user support in intelligent systems, with a focus on the importance of user control and output explanations. Understand how this research contributes to improving the interaction between people and algorithms in everyday applications.
When People and Algorithms Meet - User-Reported Problems in Intelligent Everyday Applications
Association for Computing Machinery (ACM) via YouTube
Overview
Syllabus
When people and algorithms meet: user-reported problems in intelligent everyday applications
Taught by
ACM SIGCHI