Knowledge Representation and Reasoning

Thinking requires creating and manipulating mental models that encapsulate knowledge of the world. In classical AI, these representations are symbolic and explicit, contrasting with neural networks where knowledge is buried in connection weights. Symbolic representations are interpretable, which is desirable in human-machine interaction. Just as formal mathematics addresses complex problems with precision, logic-based representations enable succinct and unambiguous knowledge representation.

The course covers propositional and first-order logic, tractable subsets including Horn clause and description logics, event calculus for time and change, and epistemic logic for agent knowledge in multi-agent systems. It also addresses inheritance with taxonomies and default reasoning under incomplete information.

Course Overview

  • Introduction. History and Philosophy.
  • Symbolic Reasoning. Truth, Logic, and Provability.
  • Propositional Logic. Direct Proofs. The Tableau Method.
  • First Order Logic. Universal Instantiation. The Unification Algorithm.
  • Forward and Backward Chaining. The Resolution Refutation Method.
  • Horn Clauses and Logic Programming. Prolog.
  • Rule Based Systems. The OPS5 Language. The Rete Algorithm.
  • Representation in First Order Logic. Conceptual Dependency.
  • Frames. Description Logics and the Web Ontology Language
  • Taxonomies and Inheritance. Default Reasoning.
  • Circumscription. Auto-epistemic Reasoning. Event Calculus
  • Epistemic Logic. Knowledge and Belief
  • Deepak Khemani. A First Course in Artificial Intelligence, McGraw Hill Education (India),  2013.
  • Ronald J Brachman, Hector J Levesque: Knowledge Representation and Reasoning,  Morgan Kaufmann, 2004.

Additional Readings

  • Schank, Roger C, Robert P Abelson: Scripts, Plans, Goals, and Understanding: An Inquiry  into Human Knowledge Structures. Hillsdale, NJ: Lawrence Erlbaum, 1977.
  • R C Schank and C K Riesbeck: Inside Computer Understanding: Five Programs Plus  Miniatures, Lawrence Erlbaum, 1981.
  • John F Sowa: Conceptual Structures: Information Processing in Mind and Machine,  Addison–Wesley Publishing Company, Reading Massachusetts, 1984.
  • Murray Shanahan: A Circumscriptive Calculus of Events. Artif. Intell. 77(2), pp. 249-284,  1995.
  • John F Sowa: Knowledge Representation: Logical, Philosophical, and Computational  Foundations, Brooks/Cole, Thomson Learning, 2000.
  • Ronald Fagin, Joseph Y Halpern, Yoram Moses, and Moshe Vardi. Reasoning About  Knowledge. MIT press, 2004.
  • Grigoris Antoniou and Frank van Harmelen, A Semantic Web Primer, 2nd Ed, MIT Press,  2008.
Top