|
|
|
 |
Advanced concepts in knowledge representation |
| |
|
15h L / 9h R / 1 OE / 2 ECTS credits / IIC_CRC |
| |
|
Abolfazl FATHOLAHZADEH |
| |
|
The aim of this module is to present and compare, with respect to their power of expression and limitations, the main methods used for knowledge representation, together with the processing and reasoning mechanisms associated with them.
Efficient representation of low-level data
Low-level data: strings, atoms. Trie, reduced trie, reduced double-trie. Finite-state automata, transducers. Combining finite state automata with decision-trees. Application to the implementation of large-scale dictionaries.
|
| |
|
Efficient representation of low-level data |
| |
|
Low-level data: strings, atoms. Trie, reduced trie, reduced double-trie. Finite-state automata, transducers. Combining finite state automata with decision-trees. Application to the implementation of large-scale dictionaries.
|
| |
|
Constraint satisfaction problems |
| |
|
Definition and formalization. Representing the constraints. Constraints graphs. Deducing new constraints. Spatial and temporal constraints. Search algorithms: FC, FC+AC, FC-CBJ. Application to task planning.
|
| |
|
Representing hard problems |
| |
|
Introduction. Limitation of Korf’s method. RWM’s benefits. Uniform representation of problems, goals and relevant, plausible or irrelevant moves. Representing the resolution strategy. Generating macros and subproblems. Domain dependent knowledge base. RWM’s architecture and resolution trace sample. Time complexity of refinement and macros generation. Applications to mod-n, Pyramix, Trillion, Magic cube, etc. RWM + TSG system and its application to automated translation.
|
| |
|
Representing fuzzy knowledge |
| |
|
Set-theoretic operations and convex fuzzy sets – Fuzzy structured sets, measures of fuzziness – Extension: principle, operations – Possibility and probability, fuzzy distribution.
Fuzzy database and fuzzy queries – Inference in the fuzzy rule-based systems – Soft constraint satisfaction problems.
|
| |
|
Learning |
| |
|
Concepts, description language, generalization language, representation samples. Generalization space: explanation tree, expression graph. Inductive learning: inductive substitution, choosing the right generalization space, induction procedure. Complexity and efficiency of inductive learning. Applications to common-sense reasoning, geometrical reasoning, medical reasoning.
Learning by examples: generalization by search.
|
| |
| |
| |
|
References
S. Russel, P. Norvig, Artificial Intelligence – A modern approach, Prentice Hall, 2002.
P. Norvig, Paradigms of Artificial Intelligence Programming: Case Studies in, Common Lisp, Morgan Kaufmann, 1992.
E. Davis, Representations of common sense knowledge,.Morgan Kaufmann, 1990. |
| |
| |
| Last update 06/07/2007 by Cl.M. |
|
