Modeling question asking using neural program generation
This addresses the challenge of enabling AI systems to ask more human-like and creative questions, which is incremental as it builds on existing neuro-symbolic and reinforcement learning methods.
The authors tackled the problem of modeling human question asking, which is richer and more creative than current AI systems, by proposing a neuro-symbolic framework that represents questions as formal programs and generates them with a deep neural network; they showed it can predict human questions in unconstrained settings and generated creative questions without supervised data.
People ask questions that are far richer, more informative, and more creative than current AI systems. We propose a neuro-symbolic framework for modeling human question asking, which represents questions as formal programs and generates programs with an encoder-decoder based deep neural network. From extensive experiments using an information-search game, we show that our method can predict which questions humans are likely to ask in unconstrained settings. We also propose a novel grammar-based question generation framework trained with reinforcement learning, which is able to generate creative questions without supervised human data.