This groundbreaking research explores how humans make decisions when risks are conveyed through text descriptions rather than numerical probabilities.

• First large-scale dataset of binary choices between textually described lotteries
• Evaluates multiple computational approaches to predict human choices
• Bridges the gap between theoretical decision science and real-world decision contexts
• Provides insights for developing AI systems that can better understand human risk perception

Medical Relevance: Understanding how patients interpret textual descriptions of treatment risks and benefits could significantly improve medical communication, informed consent processes, and shared decision-making in healthcare settings.

Predicting Human Choice Between Textually Described Lotteries