Introspective Learning and Reasoning
Introspective reasoning is a form of metareasoning in which an agent reasons about the mental objects and actions involved in its own internal reasoning processes. Introspective reasoning may guide adjustments of domain-level reasoning processes, based on internal performance goals and self-knowledge about the agent’s reasoning capabilities. Introspective learning is learning an agent performs, using introspective reasoning, to improve the future performance of its reasoning processes. The study of introspective reasoning and learning focuses on how intelligent systems – human or machine – can monitor and understand their own reasoning processes, and can exploit the resulting self-awareness to improve the performance of those processes.
Metareasoning has been studied in many disciplines, including psychology, education, computer science, and artificial intelligence. As Herbert Simon...
- Cox, M. T., & Raja, A. (Eds.). (2011). Metareasoning: Thinking about thinking. Cambridge, MA: MIT Press.Google Scholar
- Flavell, J. H. (1987). Speculation about the nature and development of metacognition. In F. Weinert & R. Kluwe (Eds.), Metacognition, motivation, and understanding (pp. 21–29). Hillsdale: Lawrence Erlbaum.Google Scholar
- Ram, A., & Leake, D. (Eds.). (1995). Learning, Goals, and Learning Goals. Goal-driven learning (pp. 1–37). Cambridge: MIT Press.Google Scholar