A Survey of Human‐Centered Evaluations in Human‐Centered Machine Learning

Visual analytics systems integrate interactive visualizations and machine learning to enable expert users solve complex analysis tasks. Applications combine techniques from various fields of research are consequently not trivial evaluate. The result is a lack structure comparability between evaluations. In this survey, we provide comprehensive overview evaluations in the field human-centered learning. We particularly focus on human-related factors that influence trust, interpretability, explainability. analyze presented papers top conferences journals information visualization human-computer interaction systematic review their setup findings. From distill design dimensions for structured evaluations, identify evaluation gaps, derive future opportunities.