• to introduce the symbolic approach to Artificial Intelligence and Cognitive Science;
• to learn the basic concepts within the symbolic approach and the basic knowl-edge representation schemas;
• to get an idea about cognitive architectures and particular cognitive models;
• to get some experience in representing knowledge in various domains.

• demonstrate familiarity with the central concepts and ideas of the symbolic approach;
• present arguments and evidence coherently in verbal and written form;
• debate and discuss theoretical issues and concepts;
• demonstrate a detailed and critical understanding of research papers - both theoretical and experimental; identify and explain significant developments and contradictions;
• select an appropriate knowledge representation schema and use it to represent the knowledge in a particular (not very complex) domain;
• critically evaluate particular symbolic cognitive models;
• argue about the advantages and limitations of the symbolic approach.

• lectures focusing on key ideas and problems;
• guided reading;
• tutor-led group discussions;
• exercises with knowledge representation techniques.

• hands-on experience with different representation techniques;
• discussions on particular topics.

Foundations of Cognitive Science, MIT Press, 1989.

[CC] Pylyshin, Z.,

Computation and Cognition: Toward a Foundation of Cognitive Science, MIT Press, Cambridge, MA, 1984.

[HAND] Barr, Feigenbaum (eds.),

Handbook of Artificial Intelligence, vol. 1-3.

[APPR] Ringland, D.,

Approaches to Knowledge Representation, John Wiley, New York, 1988.

[FK] Delgrange, Mylopolous,

Knowledge Representation: Features of Knowledge, In: Bibel, Jorrand (eds.) Fundamentals of Artificial Intelligence. Springer, Berlin, 1987.

[KS] Galambos, Abelson, Black,

Knowledge Structures, Lawrence Erlbaum, Hillsdale, NJ, 1986.

[KS] Galambos, Abelson, Black,

Knowledge Structures.

[ACT*] Anderson,

The Architecture of Cognition, Harvard Univ. Press, Cambridge, MA, 1983.

[SOAR] Newell,

Unified Theories of Cognition, Harvard Univ. Press, Cambridge, MA, 1990.

• introduction - basic ideas, methodology of symbolic Modelling;
• knowledge representation schemas - introduction to the basic knowledge rep-resentation techniques;
• cognitive architectures - overview of some symbolic cognitive architectures;
• cognitive models - providing examples of symbolic models of cognitive phe-nomena;
• concluding discussions - pros and cons, open problems, relations to the con-nectionist approach to AI.

Introduction

Topic 1: Basic concepts: Model, Modelling, Theory, Approach, Symbol, Physical Symbol System Hypothesis
Seminar: Discussion: What is a symbol?

Knowledge Representation

Topic 2: Informal introduction: historical lines of development of the idea of knowledge representation, types of information in human memory, what is representation and description, knowledge use (retrieval, reasoning, acquisition). Informal introduction of the main knowledge representation schemes by presenting examples in various domains.
Seminar: Discussion: Why do we need representation? Do we need just knowledge?

Topic 3: Formal introduction: Formal theory. Knowledge representation formalism vs. schema: syntactic rules, semantic theory, inference rules, axioms and alphabet vs. data structures and procedural operations over them. Organizational aspects. Basic operations (matching, marker passing, subsumption). Control.
Seminar: Discussion: Formalism vs. schema: similarities, differences, relations.

Required readings:

• Barr, Feigenbaum (eds.) Handbook of Artificial Intelligence, vol. 1, Chapter 3.
• Ringland, Duce - Approaches to Knowledge Representation. John Wiley, NY, 19.

Topic 4: Logics - Propositional Calculus and First-order Logic, idea for modal logic, higher-order logic, and nonmonotonic logic.
Seminar: Developing representation examples.

Required readings:

• Barr, Feigenbaum (eds.) Handbook of Artificial Intelligence, vol. 1, Chapter 3.
• Ringland, Duce - Approaches to Knowledge Representation. John Wiley, NY, 1988.

Topic 5: Production Systems
Seminar: Developing representation examples.

Required readings:

• Barr, Feigenbaum (eds.) Handbook of Artificial Intelligence, vol. 1, Chapter 3.
• Ringland, Duce - Approaches to Knowledge Representation. John Wiley, NY, 1988.

Topic 6: Semantic Networks
Seminar: Developing representation examples.

Required readings:

• Barr, Feigenbaum (eds.) Handbook of Artificial Intelligence, vol. 1, Chapter 3.
• Ringland, Duce - Approaches to Knowledge Representation. John Wiley, NY, 1988.

Topic 7: Frames
Seminar: Developing representation examples.

Required readings:

• Barr, Feigenbaum (eds.) Handbook of Artificial Intelligence, vol. 1, Chapter 3.
• Ringland, Duce - Approaches to Knowledge Representation. John Wiley, NY, 1988.

Topic 8:Properties of real-world knowledge and how to deal with them: incom-pleteness, nonmonotonity, inconsistency, inaccuracy, relativity, uncer-tainty, imprecision, etc.
Seminar: Discussion: Comparative analysis of the knowledge representation schemas.

Required reading:

• Delgrange, Mylopolous - Knowledge Representation: Features of Knowledge. In: Bibel, Jorrand (eds.) Fundamentals of Artificial Intelligence. Springer, Berlin, 1987.

Cognitive Architectures

Topic 9: Cognitive architectures, computer architectures, symbolic architectures, functional requirements.
Seminar: Discussion: Modularity of mind vs. Unified Architectures.

Required reading:

• Posner (ed.) Foundations of Cognitive Science. MIT Press, 1989 - Chapter 3.

Additional readings:

• Posner (ed.) Foundations of Cognitive Science. MIT Press, 1989 - Chapter 2.
• Pylyshyn, Z. Computation and Cognition: Toward a Foundation of Cognitive Science. MIT Press, Cambridge, MA, 1984.

Topic 10: Production systems architectures: ACT* and SOAR.
Seminar: Discussion: Pros and Cons of ACT* and SOAR.

Required reading:

• Posner (ed.) Foundations of Cognitive Science. MIT Press, 1989 - Chapter 3.

Additional readings:

• Anderson - The Architecture of Cognition. Harvard Univ. Pr., Cambridge, MA, 1983.
• Newell - Unified Theories of Cognition. Harvard Univ. Press, Cambridge, MA, 1990.
• Van Lehn (1989) Architectures for Intelligence.

Topic 11: Schema-based architectures.
Seminar: Discussion: Pros and Cons of Schema-based Architectures.

Cognitive Models

Topic 12: Problem solving (PS): Knowledge-lean PS (state space, search), knowl-edge-rich PS (schema-based models).
Seminar: Discussion: Experts vs. Novices.

Required reading:

• Anderson, J. (1990) Cognitive Psychology and its Implications. 3rd edition, Hillsdale, NJ: Erlbaum.

Topic 13:Vision - David Marr's approach to human vision.
Seminar: Discussion: Top-down vs. bottom-up approaches to vision.

Required reading:

• Anderson, J. (1990) Cognitive Psychology and its Implications. 3rd edition, Hillsdale, NJ: Erlbaum.

Topic 14:Learning models - concept formation, skill acquisition, inductive and analogical reasoning.
Seminar: Discussion: Where to start from?

Required reading:

• Anderson, J. (1990) Cognitive Psychology and its Implications. 3rd edition, Hillsdale, NJ: Erlbaum.

Concluding Discussion

Topic 15: Cognitive Representation of Emotions and Social Interaction. Verification of symbolic models, pros and cons of the symbolic approach, the symbolic/connectionist debate.
Seminar: Discussion: Do we need symbols and symbolic models?

Required reading:

• Castelfranchi, C., Conte, R., Miceli, M., Poggi, I. Emotions and Goals. In: Perspectives on Cognitive Science, vol. 2, NBU Press.

Additional reading:

• Conte, R. Castelfranchi, C. Cognitive and Social Action. UCL Press, 1995, Chapters 1-3.

• Homework assignment 1: Two different sample domains represented by at least three different representation schemas.
• Homework assignment 2: Comparative analysis of the knowledge representation schemas - a 1000 words written paper.
• Homework assignment 3: To read and comment a research paper (classic or modern) writing a 3000 words paper and defending it before the group.
• Multiple choice test.

Grading procedure:

• 15% assignment 1
• 15% assignment 2
• 25% long paper
• 25% public defense
• 20% multiple choice test