Lecture: Knowledge Acquisition and Machine Learning

Katharina Morik

The lecture consists of five blocks. At the end of each block, a session is devoted to discussion.

1. Early approaches

   early expert systems: from MYCIN to NEOMYCIN

   early knowledge acquisition (KA) systems - the transfer view (OPAL, MOLE, AQUINAS)

2. KADS

   knowledge eliciation

   modeling expertise

   applications

3. early machine learning (ML) approaches - automatic KA?

   ID3

   conceptual clustering - star method

4. current KA techniques

   OPAL to Protege

5. current ML approaches

   FOIL

   conceptual clustering - UNIMEM

   inductive logic programming, theta subsumption

   case-based reasoning

6. integration of KA and ML

   MOBAL - a KA workbench including ML tools

   knowledge maintenance

   applications of MOBAL