IA301 - Logique et IA Symbolique - Logics and Symbolic AI
Objective of the course:
Providing the bases of symbolic artificial intelligence.
Detailing a few selected advanced topics.
Acquiring skills to
understand different kinds of logic families,
formulate reasoning in such formal languages, and
manipulate tools to represent knowledge and its adaptation to imprecise and incomplete domains through the use of Protegé.
Prerequisite: Basic knowledge in computer sciences and algebra.
Course content:
Introduction
Reminder on bases on logics (syntax, semantics...) and overview of several logics (propositional, first order, modal, fuzzy...)
Some typical examples in AI: revision, merging, abduction, with illustrations on preference modeling and image understanding
Symbolic learning: formal concept analysis, decision trees
Ontologies, description logics, and knowledge graphs
Practical work on ontology engineering and design. Building your own ontologies using Protegé for real life problems (in particular related to environment)
Evaluation:
written exam (50%) -
Authorized document: 2 A4 sheets (4 pages) of notes you will have prepared in advance
reports handed in after the practical work, which will require to create a couple of ontologies as part of a decision support system of a freely elected domain problem (50%)