A Review of User Interface Design for Interactive Machine Learning

Explore a comprehensive review of user interface design for Interactive Machine Learning (IML) in this 21-minute conference talk from the 24th International Conference on Intelligent User Interfaces. Delve into the concept of IML, its importance, and the challenges it faces. Examine the structural breakdown of IML systems, including sample review, model inspection, and task overview. Discover standardized workflows and model steering techniques. Learn about emergent solution principles for building effective IML interfaces, contextualized within broader human-computer interaction literature. Gain insights into the co-adaptive nature of IML and its potential to empower non-experts in developing complex data-driven applications. Understand the key research areas crucial for advancing non-expert IML applications and enhancing human-machine collaboration in data analysis and pattern recognition.

Introduction
What is Interactive Machine Learning
Why do we care
Challenges
What has been done
What challenges remain
Topdown approach
Structural breakdown
Sample review
Model inspection
Task overview
Standardized workflow
Model steering
emergent solution principles
time for questions
methodology
workflow
final question