Mining HCI Data for Theory of Mind Induction

Abstract

Human-computer interaction (HCI) results in enormous amounts of data-bearing potentials for understanding a human user’s intentions, goals, and desires. Knowing what users want and need is a key to intelligent system assistance. The theory of mind concept known from studies in animal behavior is adopted and adapted for expressive user modeling. Theories of mind are hypothetical user models representing, to some extent, a human user’s thoughts. A theory of mind may even reveal tacit knowledge. In this way, user modeling becomes knowledge discovery going beyond the human’s knowledge and covering domain-specific insights. Theories of mind are induced by mining HCI data. Data mining turns out to be inductive modeling. Intelligent assistant systems inductively modeling a human user’s intentions, goals, and the like, as well as domain knowledge are, by nature, learning systems. To cope with the risk of getting it wrong, learning systems are equipped with the skill of reflection.