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

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. INTRODUCTIONInterpretable machine learning (ML) has attracted significant interest as ML models are used to support human decision making in societally critical domains such as justice systems and healthcare [13,21,41]. In these domains, full automation is often not desired and humans are the final decision makers for legal and ethical reasons. In fact, the Wisconsin Supreme Court ruled that "a COMPAS risk assessment should not be used to determine the severity of a sentence or whether an offender is incarcerated", but does not eliminate the use of ML models if "judges be made aware of the limitations of risk assessment tools" [40,54]. Therefore, it is crucial to enhance human performance with the assistance of machine learning models, e.g., by explaining the recommended decisions.However, recent human-subject studies tend to show limited effectiveness of explanations in improving human performance [7,23,34,62]. For instance, Lai and Tan [34] show that explanations alone only slightly improve human

Are Explanations Helpful? A Comparative Study of the Effects of Explanations in AI-Assisted Decision-Making

“…Lai and Tan [42] deception detection feature contribution N/A N/A ✓? Lai et al [41] deception detection feature contribution N/A N/A ✓? Cai et al [13] drawing recognition example-based mixed results N/A N/A Yang et al [69] leaf classification example-based N/A N/A ✓ Note: "N/A" means the study does not examine the desideratum.…”

“…To answer these questions, researchers have been advocating for moving beyond defining what constitutes a "good" explanation using model designer's intuition but actually examining how useful an explanation is with human users [24,59]. In responding to this call, there is recently a growing line of literature on empirically evaluating the effectiveness of XAI methods (e.g., [14,16,41,69,71]). Yet, principles required for an explanation to be considered helpful in AI-assisted decision making, arguably, still remain to be articulated and comprehensively assessed.…”

Intelligent Decision Assistance Versus Automated Decision-Making: Enhancing Knowledge Workers Through Explainable Artificial Intelligence

“…For example, if two XAI techniques were used and compared as separate experimental treatments we added two entries into our database. In total, we identified five articles and 12 experiments [40,56,57,58,59]. In the following, we describe the studies and their results with regard to IDA in detail.…”

Decision-Focused Summarization

“…Another related direction is identifying helpful sentences in product reviews (Gamzu et al, 2021). It is useful to highlight our motivation in support decision making in challenging tasks towards effective human-AI collaboration (Green and Chen, 2019;Lai et al, 2020;Lai and Tan, 2019). Unlike tasks such as textual entailment where models aim to emulate human intelligence, forecasting future outcomes, such as stock markets (Xing et al, 2018) and message popularity (Tan et al, 2014), is challenging both for humans and for machines.…”