Course content

Main characteristics of a knowledge representation language will be studied. Various KR paradigms will be compared with respect to their characteristics. Representation languages will be related to the underlying inference methods, and to the syntactical, semantical, and pragmatical aspects of computational representations. Advantages and disadvantages of each paradigm will be analysed.
A part of the course addresses case-based reasoning (CBR). CBR has its roots in the subfield of cognitive psychology that studies episodic memories in humans. CBR involves representation of situation-specific experiences in computers and the reuse of them in order to understand/solve a new problem.
The difference between symbolic AI and subsymbolic AI regarding the notion of representation will be discussed, and suggestions for reconciliation will be presented.
The exercises include the design and implementation of systems that use different types of representations and the corresponding reasoning mechanisms.

Learning outcome

To be able to analyse different types of representation paradigms, to explain advantages and disadvantages of each type, to choose the right type of language in a given problem. To be able to model the knowledge within a given domain. To be able to implement aknowledge-based system with a suitable representation language and inference mechanism.

Learning methods and activities

Lectures, guided colloquia, self study, and exercises. Portfolio evaluation is the basis for the grade in the course. The portfolio includes a final oral exam (80%) and a term paper (20%). The results for the parts are given in %-scores, while the entire portfolio is assigned a letter grade. There will be a set of assignments that must be approved in order to take the exam.

Compulsory assignments

• Exercises

Recommended previous knowledge

TDT4136 Logic and Reasoning Systems or equivalent.

Course materials

Textbook and a set of papers.

Credit reductions