"Why is 'Chicago' deceptive?" Towards Building Model-Driven Tutorials for Humans

TL;DR

This paper investigates model-driven tutorials to improve human understanding of machine learning models, demonstrating that tutorials based on simple models enhance human decision-making even when complex models perform better.

Contribution

It introduces a novel approach of using model-driven tutorials with explanations from simple models to aid human understanding of complex machine learning models.

Findings

Tutorials improve human performance in deception detection tasks.

Simple model explanations are more helpful to humans than complex model explanations.

Tutorials benefit human decision-making both with and without real-time model assistance.

Abstract

To support human decision making with machine learning models, we often need to elucidate patterns embedded in the models that are unsalient, unknown, or counterintuitive to humans. While existing approaches focus on explaining machine predictions with real-time assistance, we explore model-driven tutorials to help humans understand these patterns in a training phase. We consider both tutorials with guidelines from scientific papers, analogous to current practices of science communication, and automatically selected examples from training data with explanations. We use deceptive review detection as a testbed and conduct large-scale, randomized human-subject experiments to examine the effectiveness of such tutorials. We find that tutorials indeed improve human performance, with and without real-time assistance. In particular, although deep learning provides superior predictive performance than simple models, tutorials and explanations from simple models are more useful to humans. Our work suggests future directions for human-centered tutorials and explanations towards a synergy between humans and AI.