Sibyl: Understanding and Addressing the Usability Challenges of Machine Learning In High-Stakes Decision Making

TL;DR

This paper introduces Sibyl, a visual analytics tool designed to improve the interpretability and usability of machine learning models in high-stakes child welfare screening, addressing trust and ethical challenges faced by domain experts.

Contribution

The paper presents the development and evaluation of Sibyl, a novel visualization tool that enhances understanding of ML predictions for non-expert users in high-stakes decision-making.

Findings

Sibyl increased interpretability for child welfare screeners.

User studies showed improved trust and decision quality.

Provided design guidelines for domain-specific ML visualization tools.

Abstract

Machine learning (ML) is being applied to a diverse and ever-growing set of domains. In many cases, domain experts - who often have no expertise in ML or data science - are asked to use ML predictions to make high-stakes decisions. Multiple ML usability challenges can appear as result, such as lack of user trust in the model, inability to reconcile human-ML disagreement, and ethical concerns about oversimplification of complex problems to a single algorithm output. In this paper, we investigate the ML usability challenges that present in the domain of child welfare screening through a series of collaborations with child welfare screeners. Following the iterative design process between the ML scientists, visualization researchers, and domain experts (child screeners), we first identified four key ML challenges and honed in on one promising explainable ML technique to address them (local factor contributions). Then we implemented and evaluated our visual analytics tool, Sibyl, to increase the interpretability and interactivity of local factor contributions. The effectiveness of our tool is demonstrated by two formal user studies with 12 non-expert participants and 13 expert participants respectively. Valuable feedback was collected, from which we composed a list of design implications as a useful guideline for researchers who aim to develop an interpretable and interactive visualization tool for ML prediction models deployed for child welfare screeners and other similar domain experts.