Cognative Systems

Intro to KBAI

Characteristics of AI Problems

  • Knowledge arrives incrementally
  • Problems exhibit recurring patterns
  • Problems have multiple levels of granularity
  • Problems are computationall intractable
  • The world is dynamic, but knowledge is static
  • The world is open-ended, but knowledge is finite

Characteristics of AI Agents

  • Limited computing power
  • Limited sensors
  • Limited attention
  • Computational logic is fundamentally deductive
  • Agent knowledge is incomplete relative to the world

Knowledge Representations

Problem Solving Techniques


Fundamental Processes of KBAI Cognative Systems

  • "Cognative" means dealing with human-like intelligence
  • Cognative system and KBAI-agent are used interchangably
  • Deliberation, Meta-Cognition and Reaction make the three-layer system


Reasoning about the world around us.

  • Reasoning
  • Learning
  • Memory: Knowledge gained from learning needs to be stored

think of being in traffic and wanting to change lanes. Requires planning


Reasoning about deliberation or reaction.

Lane-change results in another car honking at you. You need to reason about the cause of the sub-optimal result.


  • A direct mapping from perception to action

Kinds of KBAI

  • Think/Act

  • Optimally/Like Humans

  • Think like humans: Semantic Web

  • Think optimally: Machine Learning

  • Act like humans: Improvisational Robotics

  • Act optimally: Airplane Auto-pilot

Principles of KBAI

  • KBAI agents represent and organize knowledge into knowledge structures to guide and support reasoning
  • Learning in KBAI agents is often incremental
  • Reasoning in KBAI agents is top-down as well as bottom-up
  • KBAI agents match methods to tasks
  • KBAI agents use heuristics to find solutions that are good enough, though not necessarily optimal
  • KBAI agenbts make use of recurring patterns in the problems they solve
  • The architecture of KBAI agents enables reasoning, learning and memory to support and constrain each other

  1. Advanced_Topics
  2. Analogical_Reasoning
  3. Common_Sense_Reasoning
  4. Design_and_Creativity
  5. Evaluation
  6. Fundamentals
  7. Learning
  8. Metacognition
  9. Planning
  10. Visuospatial_Reasoning